JP2959831B2 - Image data encoding device - Google Patents

Description

Description: BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoding device and an encoding method for performing high-compression encoding of image data.

[Conventional technology]

When an image signal captured by a solid-state imaging device represented by a CCD or the like is stored as digital data in a storage device such as a memory card, a magnetic disk, or a magnetic tape, the amount of data is enormous. In order to record an image in a limited storage capacity range, it is necessary to perform some highly efficient compression on the obtained image signal data. Further, in a digital electronic still camera or the like, a photographed image is stored as digital data in a recording medium such as a memory card, a magnetic disk or a magnetic tape instead of a silver halide film. The number of images that can be recorded on the magnetic tape of the volume is defined,
The recording of the specified number of images must be guaranteed, and the time required for data recording / reproducing processing must be short and constant.

Similarly, digital VTRs (video tape recorders),
Even when recording a moving image in a digital moving image file or the like, a predetermined amount of frames must be recorded without being affected by the data amount of the image per frame. That is, whether it is a still image or a moving image, it is necessary to reliably record the required number of frames, and the time required for data recording / reproduction processing is short, and
Must be constant.

Therefore, for that purpose, it is necessary to compress the image data so that the data amount per frame falls within the specified capacity.

As a high-efficiency image data compression method for that purpose, an encoding method combining orthogonal transform encoding and variable length encoding is widely known.

As a typical example, there is a method being studied in international standardization of still image coding.

 The outline of this method will be described below.

First, the image data is divided into blocks of a predetermined size, and a two-dimensional DC
Performs T (discrete cosine transform). Next, linear quantization according to each frequency component is performed, and Huffman coding is performed on the quantized value as variable length coding. At this time, the difference value between the DC component and the DC component of a nearby block is Huffman-coded. The AC component scans from a low frequency component to a high frequency component called a zigzag scan, and performs two-dimensional Huffman coding from the continuous number of invalid (zero value) components and the value of the valid component that follows. . The above is the basic part of this method.

With only this basic part, the code amount is not constant for each image because Huffman coding, which is variable length coding, is used. Therefore, the following method has been proposed as a method of controlling the code amount.

First, at the same time as the processing of the basic portion is performed, the total code amount of the generated entire screen is obtained. From the total code amount and the target code amount, an optimal quantization width for approaching the target code amount for the DCT coefficient is predicted. Next, processing after the quantization of the basic portion is repeated using this quantization width.
Then, from the total code amount generated this time, the total code amount generated last time, and the target code amount, an optimum quantization width for approaching the target code amount again is predicted. Then, the predicted quantization width is sufficiently close to the previous quantization width, and
If the total code amount generated this time is smaller than the target code amount, the process ends and the code is output. Otherwise, the process is repeated using the new quantization width.

For the quantization width, prepare a quantization matrix that is a basic form representing the relative quantization characteristics for each frequency component,
The required quantization width is obtained by multiplying this basic quantization matrix by the quantization coefficient α.

Specifically, first, the quantization in the above-described basic portion is performed using the quantization width obtained using the standard quantization coefficient α, and this is subjected to variable-length coding. The information is compared with a preset target total code amount value which is a limit to be stored, and when the total code amount reaches the target total code amount, a final encoding process is performed using the quantization width, When the total code amount is within the target total code amount, the final encoding process is performed using the quantization width, and when the total code amount does not fall within the target total code amount, the generated total code amount and the target total code amount are set. From, for example, the Newton-Raphson method (Newton-Raphso method)
n iteration) to find the optimal quantization coefficient α to approach the target total code amount, and calculate a more optimized quantization width from the obtained quantization coefficient and quantization matrix. Then, the final encoding process is performed by using this.

 The quantization width is changed in such a manner.

The above operation will be specifically described with reference to FIG.
First, as shown in (a), one frame of image data (one frame of an image presented in the
A block of a predetermined size (for example, 8 × 576 pixels)
Divided into blocks A, B, C...
As shown in (2), two-dimensional DCT (discrete cosine transform) is performed as an orthogonal transform for each of the divided blocks, and an 8 × 8
Are sequentially stored on the matrix. When viewed on a two-dimensional plane, the image data has a spatial frequency which is frequency information based on the distribution of density information.

Therefore, by performing the above DCT, the image data is converted into a DC component DC and an AC component AC, and data indicating the value of the DC component DC at the origin position (0, 0 position) on the 8 × 8 matrix, and At the 0,7 position, data indicating the maximum frequency value of the AC component AC in the horizontal axis direction, and at the 7,0 position, data indicating the frequency value of the maximum AC component AC in the vertical axis direction, Data indicating the maximum frequency value of the AC component AC in the diagonal direction is stored at each of the seven positions, and frequency data in the direction associated with each coordinate position appears at the intermediate position, with frequency higher than the origin side. Will be stored.

Next, the stored data at each coordinate position in this matrix is divided by the quantization width for each frequency component to perform linear quantization according to each frequency component (c). Huffman coding is performed as variable length coding. At this time, for the DC component DC, the difference value from the DC component of the neighboring block is represented by a group number (number of additional bits) and additional bits, the group number is Huffman-coded, and the obtained code word and additional bits are combined. As encoded data (d1, d2, e1, e2).

Regarding the AC component AC, coefficients that are not valid (the value is not “0”) are represented by a group number and additional bits.

Therefore, the AC component AC scans from a low frequency component to a high frequency component called zigzag scan, and the continuous number of invalid (value is “0”) components (zero run number) and the succeeding valid components The two-dimensional Huffman encoding is performed from the group number of the value and the obtained codeword and additional bits are combined to obtain encoded data.

Huffman coding is the above DC component per frame image
Centering on the peak of the occurrence frequency in each data distribution of DC and AC component AC, the more this center,
This is performed by encoding data in a form where bits are allocated so as to reduce the number of data bits and increase the number of bits as the number of data bits increases, thereby obtaining a code word.

 The above is the basic part of this method.

Since only the basic portion uses Huffman coding, which is entropy coding, the code amount is not constant for each image. Therefore, as a method of controlling the code amount, for example, the following process is performed.

First, using the provisional quantization coefficient α, while performing the processing of the basic portion with a quantization width for each frequency component obtained by multiplying a predetermined quantization matrix by this quantization coefficient α, The total code amount (total number of bits) in which the screen has occurred is obtained (g). From the total code amount, the target code amount, and the tentative quantization coefficient α used, the optimal quantization coefficient α for approaching the target code amount for the DCT coefficient is Newton-Raphson-Iteration. (Newt
on Raphson Iteration) (h). Next, using the quantized coefficient α (i), the above-described processing after quantization of the basic part is repeated. Then, from the total code amount generated this time, the total code amount generated last time, the target code amount, the quantization coefficient α used this time, and the quantization coefficient α used last time, The optimal quantization coefficient α to be approximated is predicted by Newton-Raphson iteration.

If the predicted quantization coefficient α and the previous quantization coefficient α are sufficiently close to each other, and the total code amount generated this time is smaller than the target code amount, the processing is terminated, and the current The encoded data is output and stored in the memory card (f). If not, the quantization coefficient α is changed, and the process is repeated using the new quantization coefficient α.

[Problems to be solved by the invention]

As described above, for example, in a digital electronic still camera or the like, the number of images that can be recorded on a single memory card or a magnetic disk device must be guaranteed. Therefore, image data is compressed and recorded. In view of operability, the processing time must be as short as possible and constant. It is also desired that image data can be compressed with high efficiency. These items are not limited to digital electronic still cameras, and are items required not only in other applications.

As a compression method that satisfies such a demand, there is the above-mentioned international standard proposal method. In this method, a method that combines orthogonal transform and variable-length coding for each block as exemplified in the above-described basic part uses image data. Compression can be performed with high efficiency, but the number of images that can be stored in a single recording medium such as a memory card or a magnetic disk device becomes indefinite due to the use of variable-length coding. There was a disadvantage.

Further, in the method of controlling the code amount as exemplified in the conventional example for solving this, not only the number of times of repeating the bus of the basic part of the encoding differs depending on the image, but also the processing time becomes indefinite. However, there is a drawback that generally requires a long processing time.

As a method for solving this problem, the present inventors have proposed the following method (Japanese Patent Application Nos. 1-28761 and 2-76).
137222).

The proposed method is a compression method that combines orthogonal transform and variable-length coding. In order to control the amount of generated codes, a sampled image signal stored in an image memory is converted to a plurality of adjacent pixels of a screen. After dividing each block into blocks and performing orthogonal transformation on each of the divided blocks, quantizing the transformed output with a provisional quantization width, and then performing variable length coding on the quantized output, , Calculate the generated code amount for each block and the total code amount of the entire image, and then calculate the provisional quantization width, the total generated code amount, and 1
A new quantization width is predicted from the target total code amount per image. At the same time, the assigned code amount for each block is calculated from the issued code amount for each block, the total generated code amount, and the target total code amount per image (statistical processing). First pass processing).

Then, the image signal of the image memory is divided again into blocks by using the new quantization width, orthogonal transformation, quantization, and variable length encoding are performed. For each block, the generated code amount is the allocated code amount of the block. Is exceeded, the variable-length coding is stopped halfway and the process of moving to the process of the next block is performed (coding process; second-pass process).

Thus, the code amount is controlled so that the total generated code amount of the entire image does not exceed the target total code amount.

However, although the above method can obtain very good results, at least a memory for storing the generated code amount of each block is required, so that the circuit scale becomes large, and the power consumption increases accordingly. Will be. The increase in power consumption means that for portable products where miniaturization is very important, the proportion of the battery part increases, the weight increases, and the overall size increases due to the increase in circuit scale. However, there is a problem in using the above method because it is extremely inconvenient and causes a cost increase for a low-priced product.

In recent years, there have been various user needs such as cases in which the quality of the image is emphasized even if the product is expensive, and a desire for an inexpensive product even if the image quality is somewhat sacrificed. Although it becomes a slightly complicated circuit like this, it is not only a method that can obtain high image quality, but even if the image quality is slightly sacrificed,
A method that can be realized by a simple circuit is also required. In addition, these two methods require that the compressed data be monitored and mutually compatible during reproduction.

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method of encoding image data that can be encoded within a predetermined code amount within a predetermined processing time with a simple circuit configuration without impairing image quality as much as possible. An encoding device and an encoding method are provided.

[Means for solving the problem]

To achieve the above object, the present invention is configured as follows. That is, according to the present invention, the image data is divided into blocks, each of the divided blocks is subjected to orthogonal transform in turn, and the transform output is quantized by a quantizing means, and the quantized output is sent to a variable length encoding means. In the coding apparatus, the variable length coding is performed and the data is encoded for each block. First, when encoding the block, the code amount generated by the encoding process in the block to be encoded and Based on the allocated code amount in the block and the amount of excess or deficiency of the generated code amount with respect to the allocated code amount, a predetermined code amount equal to or less than a target total code amount which is an allowable code amount per image is calculated. By dividing by the total number of blocks, a code amount obtained by adding the calculated excess / deficiency amount to the determined reference code amount per block is set as the allocated code amount of a block to be processed next. Code amount allocating means, and a truncation means for controlling the variable length coding means to terminate the variable length coding so as not to exceed the allocated code amount allocated by the code amount allocating means for each block. It comprises.

Secondly, a compression ratio specifying means for specifying a compression ratio,
According to the compression ratio designated by the compression ratio designation means, the permissible target total code amount and quantization width per screen and 1
Compression ratio selecting means for determining a reference allocation code amount per block, and, in coding the block, a code amount generated by a coding process in a block to be subjected to the coding process and an allocation code amount in the block. Based on the calculated code amount, the amount of excess or deficiency with respect to the allocated code amount is calculated, and the code amount obtained by adding the calculated excess or deficiency amount to the reference code amount per block determined by the compression ratio selecting means is calculated as follows. Code amount allocating means for obtaining the allocated code amount of the block to be processed, and the variable length code for terminating the variable length coding so as not to exceed the allocated code amount allocated by the code amount allocating unit for each block. Truncation means for controlling the coding means, coding output means for outputting as a code the transform coefficient subjected to variable-length coding by the variable-length coding means, To means, constituting a control means for controlling so as to perform quantization using the quantization width determined by the compression ratio selection means.

Further, according to the present invention, the image data is divided into blocks, and each of the divided blocks is subjected to orthogonal transform in order to decompose the image data of each block into low-frequency components to high-frequency components in order. An encoding device which quantizes an output by a quantizing means, supplies the quantized output to a variable-length encoding means, performs variable-length encoding, and encodes data for each block. On the basis of the amount of code generated by the encoding process in the block to be encoded and the amount of allocated code in the block, the amount of generated code and the amount of excess or deficiency with respect to the allocated code amount are calculated. The code amount obtained by adding the calculated excess / deficiency amount to the reference code amount per block predetermined according to the target total code amount, which is the allowable code amount per block, is then processed. Code amount allocating means for determining as the allocated code amount of the block to be processed,
With respect to the truncation means for controlling the variable length encoding means to terminate the variable length coding so as not to exceed the allocated code amount allocated by the code amount allocation means for each block, Control means for controlling to perform quantization using a predetermined quantization width,
The variable-length coding means obtains a difference between the DC component of the image data of each block obtained by the orthogonal transformation and the DC component in the preceding coding processing block, and uses the difference as a variable-length code. An encoding processing unit for a DC component to be converted, and an encoding processing unit for an AC component for performing variable-length encoding on the AC component after the processing of the encoding processing unit for the DC component is completed. The cut-off unit is configured to enable the cut-off control only for the AC component encoding processing unit.

Fourth, a code amount calculating unit for calculating the code amount of the transform coefficient of the entire image encoded by the variable length encoding unit, a code amount of the entire image calculated by the code amount calculating unit, A quantization width prediction unit that predicts an optimal quantization width from the permissible target total code amount per screen, and gives the predicted quantization width to the quantization unit; Based on the code amount generated by the encoding process in the block to be processed and the allocated code amount in the block, an excess / deficiency amount of the generated code amount with respect to the allocated code amount is calculated, and based on the target total code amount. Code amount allocating means for obtaining a code amount obtained by adding the calculated excess or deficiency amount to the determined reference code amount per block as the allocated code amount of a block to be processed next; The truncation means for controlling the variable length coding means to terminate the variable length coding so as not to exceed the allocated code amount allocated by the code amount allocating means; and A first encoding process for predicting an optimum quantization width using a predetermined tentative quantization width based on the quantity, and then performing the first encoding process at least once; Control means for controlling the second encoding process using the encoding width.

Further, according to the present invention, the image data is divided into blocks, each of the divided blocks is subjected to orthogonal transform in turn, and the transform output is quantized by a quantizing means, and the quantized output is given to a variable length coding means. Fifth, in a coding apparatus that performs variable length coding and data coding for each block, a compression ratio designating unit that designates a compression ratio and a compression ratio designated by the compression ratio designating unit. Compression ratio selection means for determining the total target code amount allowed per screen, the provisionally determined quantization width, and the reference allocation code amount per block; and the transform coded by the variable length coding means. A code amount calculation unit for calculating the code amount of the entire image of the coefficient, and an optimal quantization width is predicted from the code amount of the entire image calculated by the code amount calculation unit and the target total code amount. A quantization width prediction unit that gives the prediction quantization width to the quantization unit; and a code amount generated by an encoding process in the block to be encoded and an assigned code in the block when encoding the block. The amount of excess or deficiency of the generated code amount with respect to the allocated code amount is calculated based on the amount and the code obtained by adding the calculated excess or deficiency amount to the reference code amount per block determined by the compression ratio selecting means. Code amount allocating means for determining the amount as the allocated code amount of a block to be processed next, and the variable length coding for terminating the variable length coding so as not to exceed the allocated code amount allocated by the code amount allocating unit for each block. Optimum quantization using a provisional quantization width predetermined from the target total code amount for the truncation means for controlling the variable length coding means and the quantization means. Control means for performing at least once a first encoding process for estimating, and then performing a second encoding process using the optimum quantization width predicted by the quantization width prediction means. Are provided.

Sixth, the image data is divided into blocks, and each of the divided blocks is subjected to orthogonal transform in order to decompose the image data for each block from low frequency components to high frequency components in order. In an encoding device, the output is quantized by a quantizing unit, the quantized output is supplied to a variable-length encoding unit and subjected to variable-length encoding, and data is encoded for each block. Code amount calculating means for calculating the code amount of the transform coefficient of the entire transformed image, and the optimal quantum based on the code amount of the entire image calculated by the code amount calculating means and the permissible target total code amount per screen. A quantization width prediction unit that predicts the quantization width and provides the prediction quantization width to the quantization unit; and, in encoding the block, codes in a block to be encoded. Calculating the excess / deficiency of the generated code amount with respect to the allocated code amount based on the code amount generated by the conversion process and the allocated code amount in the block, and calculates the amount of error per block determined based on the target total code amount. A code amount allocating means for calculating a code amount obtained by adding the calculated excess / deficiency amount to the reference code amount of the block as the allocated code amount of a block to be processed next, and the quantizing means, first, from the target total code amount. A first encoding process of predicting an optimum quantization width using a predetermined temporary quantization width is performed at least once, and then an optimum quantization width predicted by the quantization width prediction unit is calculated. Control means for performing a second encoding process by using the variable-length encoding means, wherein the variable-length encoding means includes a DC component of the image data of each block obtained by the orthogonal transform. After obtaining the difference from the DC component in the preceding coding processing block and performing the variable-length coding on the difference, the coding processing unit for the DC component and the processing of the coding processing unit for the DC component are completed. A coding processing section for an AC component that performs variable-length coding on the AC component, and the coding cut-off means enables the cut-off control only for the coding processing section for the AC component.

Seventh, the image data is divided into blocks, each of the divided blocks is sequentially subjected to orthogonal transform, and then quantized, and the quantized data is subjected to variable-length coding, so that In a coding method for coding data, a first step of checking the code amount of the entire image by performing the data coding process, and optimizing the code amount of the entire image obtained in the first step A second step of predicting a quantization width required for the above, a third step of performing quantization for each of the blocks using the predicted quantization width, and a step of encoding the block. Based on the code amount generated by the encoding process in the block to be processed and the allocated code amount in the block, an excess / deficiency amount of the generated code amount with respect to the allocated code amount is calculated. A code amount obtained by adding the calculated excess / deficiency amount to a reference code amount per block predetermined according to a target total code amount which is an allowable code amount is obtained as the allocated code amount of a block to be processed next. It is characterized by comprising 4 steps and a 5th step of performing variable length coding of the block within the range of the allocated code amount for each block in the fourth step.

Eighth, the image data is divided into blocks, and each of the divided blocks is subjected to orthogonal transform in order to decompose the image data for each block from low frequency components to high frequency components in order. In a coding method of coding data for each block by quantizing the decomposed data and performing variable length coding, the block unit obtained by the orthogonal transform in the variable length coding process Of the image data of the DC component, a difference from the DC component in the preceding encoding processing block is obtained and the difference is subjected to variable-length encoding. After the encoding process for the DC component ends, the AC component A variable-length encoding of the image, and a first step of examining the code amount of the entire image necessary for optimization, and the entire image obtained in the first step. A second step of predicting a quantization width required for optimization based on the code amount of the field, a third step of performing quantization for each block using the predicted quantization width, and In encoding
Based on the code amount generated by the encoding process in the block to be encoded and the allocated code amount in the block, an excess / deficiency amount of the generated code amount with respect to the allocated code amount is calculated. A code amount obtained by adding the calculated excess or deficiency amount to a reference code amount per block determined in advance according to a target total code amount which is a code amount to be processed is obtained as the allocated code amount of a block to be processed next. Characterized by the following steps:

[Operation] In such a configuration, the image data is divided into blocks, and each of the divided blocks is subjected to orthogonal transform for each block in order to obtain a transform coefficient. After performing quantization by the quantization means,
This is variable-length coded. In the case of the first configuration, 1
By dividing a predetermined code amount equal to or smaller than a target total code amount, which is an allowable code amount per image, by the total number of blocks, the determined reference allocation code amount per block and a block allocated before this block can be obtained. The generated code amount of the encoding process in the block currently being processed, based on the allocated code amount for the currently processed block obtained by combining the generated code amount in the encoding process with the excess or deficiency with respect to the allocated code amount. Is monitored, and the variable-length coding process is performed so that the coding in the block is terminated before this exceeds the allocated code amount.

In the case of the second configuration, by selecting a compression rate, a target total code amount and a quantization width corresponding to the selected compression rate and a code amount per block are determined.
After performing orthogonal transformation by blocking image data and performing quantization with the quantization width, the reference code amount per block and the generated code amount in the coding process in a block before this block are calculated. Based on the allocated code amount for the currently processed block obtained by combining the excess and deficiency with respect to the allocated code amount, the generated code amount of the encoding process in the currently processed block is monitored. Variable length encoding is performed so that the encoding in the block is terminated before the value exceeds.

In the case of the third configuration, the image data is divided into blocks, and the divided blocks are subjected to orthogonal transform for each block in order, and the image data is decomposed into low-frequency components and high-frequency components in order. This transform output is quantized by a quantizer, and then the quantized output is supplied to a variable-length encoder to perform variable-length encoding and data encoding. Of the image data, for the DC component, a difference from the DC component in the block before this block is obtained, and the difference is subjected to variable-length coding. After the coding process for the DC component is completed, the AC component is obtained. , A reference code amount per block determined in advance according to the target total code amount, and a generated code amount in an encoding process in a block before this block. Monitoring the generated code amount of the encoding process in the block currently being processed based on the allocated code amount for the currently processed block obtained by combining the excess and deficiency with respect to the allocated code amount, and Before exceeding the amount, a variable length encoding process is performed so as to terminate the encoding in the block.

In the case of the fourth configuration, image data is divided into blocks, orthogonal transformation is performed, quantization is performed with a provisional quantization width, variable-length coding is performed, and the code amount of the entire image is calculated. Predict the optimal quantization width. Then, the orthogonally transformed transform coefficients are again quantized with a new quantization width, and a reference allocation code amount per block determined in advance according to the target total code amount, and a reference allocation code amount before this block. Based on the allocated code amount for the currently processed block obtained by combining the generated code amount in the coding process of the block with the excess or deficiency with respect to the allocated code amount. The variable length coding process is performed so as to terminate the coding in the block before the generated code amount exceeds the allocated code amount.

Further, in the case of the fifth configuration, by selecting a compression ratio, the target total code amount, the provisional quantization width, and the reference allocation code amount per block are determined according to the selected compression ratio. After performing orthogonal transformation by blocking data and performing quantization with the provisional quantization width, variable-length coding is performed, and the code amount of the entire image is calculated. After the orthogonally transformed transform coefficients are again quantized with the new quantization width obtained by the prediction, the reference allocation code amount per block and the coding in the block before this block are performed. The generated code amount of the encoding process in the block currently being processed is monitored based on the allocated code amount for the currently processed block obtained by combining the generated code amount in the processing with the excess or deficiency of the allocated code amount. And this is Performing variable length coding process to abort the coding in the block before exceeding the amount of codes.

In the case of the sixth configuration, the image data is divided into blocks, and the divided blocks are sequentially subjected to orthogonal transform for each block, and the image data for each block is sequentially converted from a low frequency component to a high frequency component. Decompose the components, quantize the transformed output by the quantizing means, and then provide the quantized output to the variable-length coding means to perform variable-length coding,
Data is encoded. Among image data of each block obtained by the orthogonal transform, a difference between a DC component and a DC component of a block before this block is obtained, and this difference is variable-length. After the coding process for the DC component is completed, the coding amount of the entire image necessary for optimization is checked for the AC component, and the optimum quantization width is thereby predicted. Quantization of each block is performed using a width, and a reference allocation code amount per block, which is determined in advance according to a target total code amount, and a generation amount in a coding process in a block before this block. The code amount, based on the allocated code amount for the currently processed block obtained by combining the excess and deficiency with respect to the allocated code amount, based on the generated code amount of the encoding process in the currently processed block. Vision, and which performs variable length coding process to abort the coding in the block before exceeding the allocated code quantity.

Further, in the seventh method, the image data is divided into blocks, and each of the divided blocks is subjected to an orthogonal transform in order, then quantized, and the quantized data is subjected to variable-length coding. In encoding data for each block, the encoding process of the data is performed to check the code amount of the entire image, and the prediction of the quantization width required for optimization is performed based on the code amount of the entire image obtained thereby. Performing quantization for each block using the predicted quantization width, and, when encoding the block, the amount of code generated by the encoding process in the block to be encoded and Of the generated code amount with respect to the allocated code amount based on the allocated code amount, and is determined in advance according to the target total code amount which is the allowable code amount per image. The code amount obtained by adding the calculated excess / deficiency amount to the reference code amount per block is determined as the allocated code amount of the next block to be processed. Perform variable length coding of blocks.

Further, in an eighth method, the image data is divided into blocks, and each of the divided blocks is subjected to orthogonal transform in order to decompose the image data for each block from low frequency components to high frequency components in order. After quantizing the component-decomposed data, the data is encoded for each block by performing variable-length encoding.
In the variable-length encoding process, of the image data in each block obtained by the orthogonal transform, a difference between the DC component and the DC component in the preceding encoding processing block is obtained, and this difference is referred to as a variable-length code. After the coding process for the DC component is completed, the AC component is subjected to variable-length coding, thereby checking the code amount of the entire image necessary for optimization, and thereby obtaining the code amount of the entire image. Based on the prediction of the quantization width required for optimization, perform quantization for each block using the predicted quantization width, and, when encoding the block, in the block to be subjected to the encoding process Calculating the excess / deficiency of the generated code amount with respect to the allocated code amount on the basis of the code amount generated by the encoding process and the allocated code amount in the block. Determined as the assigned code amount of a block to be processed next code amount in which the addition of the calculated deficiency amount to the reference code amount per block which is determined in advance according to the total code amount of interest is a code amount allowed.

As described above, the present apparatus determines the code amount per block from the target code amount, and controls the variable-length coding in block units so as to be within the desired code amount while sequentially watching the coded output. Then, an encoded output is obtained as a final output.

In addition, the present apparatus performs at least one statistical process first, optimizes the quantization width, and then performs the encoding process, so that a better result can be obtained. The amount of code of the entire image necessary to perform the coding is checked, and then the processing for performing the optimized coding based on the information obtained by the statistical processing is performed. The variable length coding is controlled on a block basis so as to be within the amount, and the coded output is obtained as the final output.

Therefore, first, the image is blocked, the quantization is performed using a quantization width determined in advance from the target code amount for the elements of the blocked image, and the transform coefficient obtained by the quantization is varied. In the long coding and the variable length coding, 1 is obtained in advance from a target code amount.
Variable-length coding of the relevant element within the range of the allocated code amount of the block, including the allocated code amount per block and the excess or shortage of the allocated code amount generated by the previous coding, and all codes of the image to be processed Since the final output is obtained by performing a process called an output process for storing the image data, the obtained image data can be encoded so as to be within a certain code amount. In the case where the number of stored images is specified, for example, a fixed number of storable images can be guaranteed for the storage capacity of the storage unit. To perform encoding. In addition, since the control of the code amount uses only the allocated code amount per block and the excess or deficiency of the allocated code amount up to the block, a large storage area is not required and a relatively simple configuration can be realized. Further, since the coding is discontinued for high frequency components, the deterioration of the image quality is relatively small.

Also, after first performing statistical processing using the provisional quantization width at least once to predict the optimal quantization width,
In the method of encoding using the optimal quantization width obtained based on this, since the processing ends in at least one fixed number of passes of the statistical processing pass and the encoding processing pass,
Encoding can be performed within a certain processing time,
Occurrence of code truncation in encoding is extremely reduced, and since the code truncation is performed on high frequency components, there is almost no deterioration in image quality.

Therefore, according to the present invention, with a simple configuration, encoding can be performed within a fixed code amount within a fixed processing time, and there is almost no deterioration in image quality. Further, according to the present invention, since the code itself is not affected, the compatibility at the time of reproduction described in the international standard scheme shown in the conventional example and the subject of the present invention can be maintained. In other words, the playback device may be common.

〔Example〕

 Hereinafter, embodiments of the present invention will be described with reference to the drawings.

First, to make the present invention easy to understand, a basic concept of the present invention will be described.

That is, in the present invention, for example, one image (screen) is divided into blocks, and the data of the constituent pixels of the blocks are orthogonally transformed in block units in order, and further quantized and compression-coded. The image data of each block is sequentially quantized by a quantization width determined in advance based on a target total code amount which is a code amount desired to be finally contained in one image, and the quantized data is encoded. The code amount obtained by this coding is determined by the reference allocation code amount per block determined in advance from the target total code amount and the coding process in the block that is in the processing order one block before this block. The code amount is set so as to be within the allocated code amount which is the allowable code amount of the block determined by the amount obtained by adding the excess or deficiency to the allocated code amount of the generated code amount. View to promoting the encoding while, EOB (E
nd Of Block), when the code amount reaches the allocated code amount, the coding of the block is terminated. However, even when the DC component exceeds the adjusted allocated code amount, the DC component is always coded. The excess or deficiency for the code amount is stored, and the process proceeds to the encoding of the next block, and in the next block, the excess or deficiency is added to the reference allocated code amount and adjusted. The variable length coding is performed within the range of the allocated code amount.

Further, as another concept, the present invention quantizes the data of the constituent pixels of each block in order in a block unit, and compresses and codes the data. The image data of each block is sequentially quantized by a quantization width determined in advance based on the target total code amount, and an optimum quantization width is predicted from the code amount, and again using the predicted optimum quantization width. ,
While quantizing the image data of each block in order,
This quantized code is coded, and the code amount obtained by this coding corresponds to the reference allocation code amount per block determined in advance from the target total code amount and the processing order immediately before this block. The code amount is monitored so as to be within the allocated code amount, which is the allowable code amount of the block, which is determined by the amount obtained by adding the excess or deficiency to the allocated code amount of the generated code amount in the corresponding encoding process in the block. When the code amount reaches the allocated code amount, including the EOB code, the coding of the block is terminated, but even if the DC component exceeds the adjusted allocated code amount, the block is always coded. The amount of excess or deficiency with respect to the allocated code amount is stored, and the process proceeds to the encoding of the next block, and the amount of excess or deficiency is adjusted by adding the excess or deficiency amount to the reference allocated code amount in the next block. The assigned code amount using the clock, in which say that the variable length coding in the range of the assigned code amount.

The former is a one-pass method, and the latter is a two-pass method. In both cases, when quantizing with a certain quantization width and performing variable-length coding, the reference allocation code amount allocated to each block is used as one reference. Is to adjust by absorbing the variation in the generated code amount in the next block. This adjusted value is used as the allocated code amount, and the coding in the block is performed within the range of the allocated code amount. However, when the allocated code amount is exceeded, the coding of the block is terminated. However, the image data is decomposed into a DC component and an AC component by orthogonal transformation, and the coding is not discontinued even if the code amount exceeds the allocated code amount for the DC component that has a large influence on the visual component. Is deducted by the excess amount, and if there is a danger of entering the AC component and exceeding the allocated code amount, the coding is terminated. When the code amount generated by encoding is less than the allocated code amount, the surplus amount is added to the allocated code amount of the next block, and the allocated code amount in the next block is increased. Concerning the discontinuation of the AC component, the fact that a higher frequency component tends to have less visual influence as the frequency becomes higher is used. In this way,
When the code amount in the immediately preceding processing block exceeds or decreases the allocated code amount, the excess or deficiency is reflected in the allocated code amount of the next block and adjusted. There is a feature.

To describe the two-pass method in detail, in the present invention, first, as the first pass processing, statistical processing is performed using a provisional quantization width obtained in advance from the target total code amount, and an optimal quantization coefficient is calculated. Predict. Next, a final encoding process is performed as a process of the second pass. In the second pass, each block is quantized by the predictive quantization coefficient and encoded, and the code amount obtained by this encoding is determined per block determined in advance based on the target total code amount. Of the reference allocation code amount and the block (a) immediately before that block
(A difference between the allocated code amount allocated to the immediately preceding block (a) and the generated code amount in the current block (a)) is added. While monitoring the code amount so as to be within the allocated code amount determined by the amount, the coding is advanced, and when the code amount including the EOB code reaches the allocated code amount, the coding of the block is terminated, but the DC component Is always encoded even if it exceeds the allocated code amount, the excess or shortage of the allocated code amount at that time is stored, and the process proceeds to the encoding of the next block.

The code amount control based on the coding termination for each block using the allocated code amount per block can be used to control the code amount of the entire image by itself, or the code at the time when the approximate code amount control is performed. It can also be used for fine adjustment when the amount exceeds the target code amount.

The case of performing encoding in one pass without performing statistical processing is the one-pass method described above. Since the encoding is completed in one pass, a very large effect that processing can be performed in a short time can be obtained. However, if the system does not always encode similar images, that is, images having the same code amount generated by encoding, very good image quality may not be obtained depending on the image.

On the other hand, when the statistical processing is used, the two-pass method is used. When this statistical processing is combined with the code amount control by block-based coding termination using the allocated code amount per block, particularly, A great effect can be obtained.
Hereinafter, this case will be described in detail as an example.

The statistical processing predicts an optimal quantization width. By using the optimized quantization width in the encoding processing, it becomes possible to approximate the target total code amount. At this point, if the code amount falls within the target total code amount, this processing alone is sufficient. However, when the upper limit of the data amount of one image is defined, let alone 1 byte,
It is not permissible for the bit to exceed the target code amount. Therefore, a processing method when overrun is required.

That is the control based on the allocated code amount for each block. This is used for fine adjustment when the code amount at the time of encoding exceeds the target total code amount (a value corresponding to the upper limit of the data amount of one image). The result of executing the encoding process at the optimum quantization width predicted in the statistical process may be referred to as “end” if not exceeded, and “post-processing” if the result is exceeded. , Encoding and post-processing, which is not only time-consuming, but also requires saving so that codes of different lengths can be distinguished between the encoding and post-processing without joining. However, since there is a problem that a large amount of memory is required, it is desired to perform fine adjustment during the encoding process.
However, dropping data indiscriminately leads to deterioration of image quality, and therefore must be avoided.

Therefore, in the present invention, visual effects are minimized by omitting high frequency components in each block. However, it is impossible to determine whether or not the code amount is exceeded until the coding is completed. Therefore, in the present invention, it is determined for each block at the time of coding.

A predetermined tentative quantization width and an allocated code amount per block are determined in advance according to the target total code amount, and the code amount of the entire image is determined as long as the code amount of each block does not exceed this. Will not exceed the code amount of this block ", and this guideline is used as a reference for monitoring as the allocated code amount for each block. Then, in the encoding process according to the present invention, in each block, the encoding is terminated in each block so as not to exceed the allocated code amount of the block.

Each block is decomposed into individual frequency components according to the spatial frequency of the image data in the block by orthogonal transformation, and the data is quantized. When the components are sequentially encoded, the encoding is performed while monitoring so as not to exceed the guideline (assigned code amount). Here, in general, an image is used in which the guideline is too large, and there is a block in which coding is completed without using all of the allocated code amount and a block in which the coding amount is insufficient. Therefore, the block that has not exceeded, that is, the block in which the code amount generated including the EOB is smaller than the allocated code amount or just the same, terminates the coding without any problem, that is, outputs the EOB. Blocks that are overrun on the way are discontinued before coding exceeds the allocated code amount, and high frequency components beyond that are cut off without coding, so that they fall within the allocated code amount. In short, the encoding of the block that ends in the middle of the block is completed before the block is over, that is, the EOB is output. Since the remainder of the allocated code amount often occurs during this halfway truncation (since variable length coding causes a change in the code length depending on the data value, there is often a surplus when truncation occurs halfway). In addition to the reference code amount of one block,
It is assumed to be the allocated code amount of the next block.

Unless the compression ratio is extremely high, this is not possible, but in the unlikely event that the assigned code amount is insufficient to encode even the DC component, only the DC component is encoded and the encoding is performed. The encoding of the block is completed, that is, the EOB is output. Then, the insufficient code amount is taken as a shortage and carried over to the next block, and the shortage is subtracted from the reference allocation code amount of one block to obtain the allocation code amount of the next block. At this time, since the EOB is one of the Huffman codes, it is necessary to control the coding to be discontinued so that the EOB and the EOB are within the allocated code amount.

In this way, for example, if most blocks end coding without having to abort, and some of the remaining parts have a very high frequency omitted, and the coding is finished, the missing information will be lost. Is extremely small, and the missing information can be limited to information of high frequency components which have little visual influence. With this method, encoding can be always completed in two steps of statistical processing and encoding processing, so that it is not necessary to repeat optimization as many times as in the past, and the encoding is performed by a relatively simple mechanism. Control the volume and without the need for a lot of working memory as before,
The total code amount can be kept within the specified value,
In addition, deterioration of image quality can be suppressed.

According to the present invention, the amount of generated code in each block of an image is different. That is, although there is a local deviation in the amount of code, a block of several blocks or several tens of blocks is considered as a microblock. The code amount pays attention to relatively small deviation. Therefore, even if the reference allocated code amount per block is fixed, the surplus amount of the generated code amount with respect to the allocated code amount is carried over to the allocated code amount of the subsequent blocks, so that the generated code amount of each block is reduced. This means that the generated code amount is controlled in the micro block while adapting to the change.

According to this principle, if blocks with a large amount of generated code are concentrated at the very beginning of the encoding, the original effect may not be obtained.However, such a case is rarely possible. It ’s less of a problem, because the effect is likely to appear in less important parts of the image,
For example, 90% of the target code amount of the entire image is divided by the number of blocks to obtain a reference allocated code amount per block, and the remaining 10% is determined as if it is a surplus of the allocated code amount in the previous block. This can be solved by adopting a method that is given at the time of encoding the first block.

 An embodiment of the present apparatus using the above principle will be described.

FIG. 1 shows an embodiment in which the image data encoding apparatus according to the present invention is applied to a digital electronic still camera.
FIG. 2 is a block diagram showing the configuration of an image data encoding device according to the present invention. The illustration and description of the mechanism of the digital electronic still camera which is not directly related to the present invention is omitted.

As shown in FIG. 1, a digital electronic still camera main body (hereinafter, referred to as an electronic camera main body) 1 includes an imaging system 40 for capturing an image, and a signal processing for performing a predetermined signal processing on an output of the imaging system 40. Circuit 60, orthogonal transform, linear quantization,
An encoding circuit 80 having a variable-length encoding function and the like for compressing and encoding the output of the signal processing circuit 60 and outputting the image data and the quantization width encoded by the encoding circuit 80 (or corresponding to the ) On a recording medium 71, a switch 30 for setting and inputting a desired data compression ratio, and a control circuit 90 for controlling the entire system.

The operation unit of the electronic camera body 1 is provided with a switch 30 for setting an image compression ratio.
Connected to 90.

The imaging system 40 includes a lens 40a for forming an optical image.
And an image sensor 40b such as a CCD. The signal processing circuit
Reference numeral 60 denotes an amplifier 60a for performing amplification and noise removal, and an A / D converter 60b for converting an analog signal to a digital signal.
A buffer memory 60c including a RAM and the like, and a process circuit 60d for performing color signal formation and the like are provided. The encoding circuit 80 includes, for example, an orthogonal transformation circuit 4 that performs orthogonal transformation such as DCT (discrete cosine transformation), a quantization circuit 6 that performs linear quantization, and a Huffman encoding circuit 8 that performs Huffman encoding as variable-length encoding.
And further, a quantization width prediction circuit 12, a code amount calculation circuit 14,
Code amount allocating circuit 20, code truncation circuit 16, and coding circuit 80
And a control circuit 18 for performing a control process inside.

The recording system 70 is a memory card 71 having a built-in interface circuit 70a and an IC memory used as a recording medium.
Consists of The memory card 71 is detachable from the electronic camera body 1. The control circuit 90 is realized by a microprocessor (MPU).

FIG. 6 is a perspective view showing the external appearance of the electronic camera body 1. The figure shows a binocular type, 48 is the LC in the operation unit
A D (liquid crystal) display 30 is a switch 30 in the operation unit, and is further provided with a tele-wide switch, a shutter operation button 50, and the like. 49 is a finder. Under the control of the control circuit 90, the LCD display 48 displays various values and states such as a shooting mode, the number of frames, date and time. In this electronic camera, the image compression ratio can be set to a desired value by operating the switch 30 provided on the operation unit of the electronic camera body 1. That is, standard plural types of compression ratio information are set in the control circuit 90 in advance, and based on the value of the number of recordable images set by the operation of the switch 30, this is set in the attached memory card (recording The compression rate to be applied is obtained from the capacity of the medium, and is converted into the value of the obtained compression rate and the value of the number of images that can be recorded on the memory card, and the LCD display 48
Is displayed. Then, when the user presses the switch 30, the control circuit 90 changes these values each time the switch 30 is pressed.

The user stops pressing the switch 30 at the desired value while watching the displayed change value, so that the control circuit 90 sets the designated compression ratio corresponding to the designated number of shootable images at that time. It has become. This is a control circuit that controls the standard total code amount per image determined according to the compression ratio.
The calculation is performed by the 90 and given to the encoding circuit 80 as target code amount setting information. When the shutter operation button 50, which is a trigger switch, is pressed, a shutter function is activated and a subject image is formed on the image sensor 40b.
Since a charge image is stored in the image sensor 40b corresponding to this image, a video signal can be obtained from the image sensor 40b by reading and controlling the charge image. These controls are also performed by the control circuit 90.
Governs.

The imaging system 40 in FIG. 1 has an imaging device 40b composed of an imaging device such as an imaging lens 40a and a CCD, and converts an optical image formed on the imaging device 40b by the imaging lens 40a into an image signal. The signal is output to the signal processing circuit 60.
The signal processing circuit 60 includes an amplifier 60a, an A / D converter 60b, a buffer memory 60c, and a process circuit 60d. The process circuit 60d converts the image signal obtained by the image pickup device 40b into Y, RY of color signals. (Hereinafter, this RY is abbreviated as Cr (chroma red)), BY (hereinafter, this BY is referred to as
Cb (Chroma Blue)), and separates each color component, and performs gamma correction and white balance processing.

The output video signal of the imaging system 40 that has been digitally converted by the A / D converter 60b is, for example, image data stored in a buffer memory 60c having a capacity of one frame, read, and provided to a process circuit 60d. , The Y component, which is a luminance signal system, and the Cr, C, which is a chroma (C; color difference signal) system
It is separated into b components. The image data stored in the buffer memory 60c is processed by a process circuit to obtain Y component data of the image signal, for example, in order to first perform statistical processing on a luminance signal, and the obtained Y data is given to the encoding circuit 80. , And Y component data, and when the process is completed, next, process the chroma Cr and Cb component data and then perform the encoding process.

The signal processing circuit 60 has a blocking function. The signal processing circuit 60 reads out from the buffer memory 60c and processes the Y.
Blocking processing for dividing the image data (for one frame or one field) for the component and the Cr and Cb components into blocks of a predetermined size can be performed. Here, the block size is 8 × 8 as an example, but this block size is not limited to 8 × 8, and the block size may be different between Y and C (chroma system).

In the present embodiment, the data of the luminance system Y is read out and divided into blocks, which are provided to a subsequent processing system to perform statistical processing on the Y component data. When the statistical processing is completed,
Next, in order to enter the statistical processing for the data of the chroma Cr and Cb components, the reading of the data of the chroma Cr and Cb components and the blocking are started. In the chroma blocking, it is assumed that all the image data of the Cr component are firstly blocked, and then the image data of the Cb component is blocked.

The encoding circuit 80 has the configuration shown in FIG. Second
In the figure, reference numeral 4 denotes an orthogonal transformation circuit which receives each piece of image data input as a block and performs two-dimensional orthogonal transformation on this image data for each block. As the orthogonal transform, cosine transform, sine transform, Fourier transform, Hadamard transform and the like can be used. By performing the orthogonal transform, image data as a transform coefficient is obtained. Since the image data as the transform coefficient is an orthogonal transform, it is data obtained by decomposing the spatial frequency of the image for each frequency component.

Numeral 6 denotes a quantization circuit, which receives image data (transformation coefficient) output from the orthogonal transformation circuit 4 and sets a predetermined quantization width for each frequency component in the first quantization, and switches to a shooting mode. Quantization of the transform coefficient is performed using the quantization width corrected by multiplying the quantization coefficient α set in advance in accordance with the quantization coefficient. In the second time, quantization is performed using the optimal quantization coefficient α determined in the previous process. There is a configuration to perform.

Reference numeral 8 denotes a variable length coding circuit. The variable length coding circuit 8 performs variable length coding on the quantized output output from the quantization circuit 6. Huffman coding, arithmetic coding and the like are used as variable length coding. In variable length coding, the code amount for each block, the code amount for the entire image, and the like change for each image. Although what kind of variable length coding is used is not directly related to the present invention, an example using Huffman coding will be described here.

The variable length encoding circuit 8 scans from the low frequency components to the high frequency components by a method called zigzag scanning in which the input quantized transform coefficients are scanned in the order shown in FIG. The data of the first DC component [DC] in the scanning order in FIG. 9 is obtained by Huffman-encoding the difference value between the DC component of the block on which the variable-length coding has been performed immediately before and output. For the AC component [AC], look at the conversion coefficients in order from the second to the 64th in the scanning order in FIG. 9, and if a conversion coefficient that is not 0 (ie, a valid coefficient) comes out, it exists immediately before that. An operation of performing two-dimensional Huffman encoding on the number of consecutive 0 (invalid) coefficients (zero run) and the value of the effective coefficient is output.

When an invalid coefficient continues from a certain coefficient to the 64th coefficient, an EOB (end of block) code indicating the end of the block is output. Also, when the censoring signal is input, the coding is terminated, and EOB is added and output. The code length generated for each code is output to the code amount calculation circuit 14.

The code amount calculation circuit 14 calculates the code amount of each of the Y, Cr, and Cb components by calculating the code length of each of the input Y, Cr, and Cb components, and calculates the code amount data of the entire image. Is output to the quantization width prediction circuit 12, and the code amount data for each color component and the code amount of the entire image are output to the code amount assignment circuit 20.

At the start of the first pass, the quantization width prediction circuit 12 receives information on the target code amount from the control circuit 18, and from this code amount information, the following equation log BR = a × log SF + b (1) where BR is The code amount (the number of bits per pixel), SF is a quantization coefficient, and a and b are constants.

The initial value of the quantization coefficient α is set by using the relationship (1) and is output to the quantization circuit 6, and prior to the start of the second pass, the code amount of the entire image input from the code amount calculation circuit 14 and 1 From the target total code amount, which is the maximum amount of data per image, the optimum quantization coefficient α for approximating the target total code amount by linear prediction using, for example, the relationship of equation (1), The prediction is performed in consideration of the actually used quantization coefficient.

Here, how to make the quantization coefficient an optimum value is an important issue, and this point will be described a little.

It is well known that when image data is pre-processed, its output is quantized, and when this quantized output is subjected to variable-length encoding, changing the quantization width of the quantization changes the amount of code generated. . This is because variable-length coding represented by Huffman coding reduces the amount of code necessary to represent data using the bias in the probability of occurrence of the data to be coded. Since "changing the quantization width" also means changing the probability of occurrence of the quantization value, it can be seen that changing the quantization width also changes the generated code amount.

By the way, even if the same encoding is performed with the same quantization width,
The generated code amount differs depending on the image data at that time. However, when the same encoding is performed by changing the quantization width for one image data, a fixed relationship is obtained between the quantization width and the generated code amount. In addition, when the relationship between the quantization width and the generated code amount is obtained for many pieces of image data, it has been found that the relationship having the highest frequency of occurrence can be obtained statistically.

Specifically, in many cases, the following relationship was obtained. That is, if the relative ratio with respect to a certain quantization width is SF, and it is represented by the number of bits (bit rate) per pixel of the generated code amount, and is represented by BR, the relationship of the above equation (1) holds.
In the equation (1), if a is the same encoding, a is substantially constant regardless of the image, and b depends on the image. The value of b has a certain distribution depending on the image, and a typical b is obtained from the occurrence frequency distribution.

Therefore, the optimum quantization coefficient can be obtained from the relationship of the expression (1).

The code amount allocating circuit 20 calculates the allocated code amount per block for each color component from the code amount of the image data for each color component input from the code amount calculation circuit 14, the code amount of the entire image, and the target total code amount. Is calculated and output to the encoding truncation circuit 16.

In this calculation method, for example, the target total code amount is proportionally distributed based on the ratio of the code amount for each color component, and the allocated code amount for each color component is further divided by the number of blocks. For example, the code amount of a certain color component is multiplied by the target total code amount, and the result is divided by the code amount of the entire image, and then further divided by the number of blocks of the color component. Is determined. As a result, the assigned code amount of each color component is
If the code amount is small in accordance with the actual code amount of the color component, the code amount is appropriately suppressed to a sufficient extent, and the color component with a large code amount is allocated a correspondingly large amount.

The code amount allocating circuit 20 has a code amount information table and a block allocated code amount data table, and rewrites the code amount information of the corresponding block position in the code amount information table with the code amount information input from the code amount calculating circuit 14. From the code amount for each color component and the code amount of the entire image input from the code amount calculation circuit 14, and the target total code amount, one of the color components is calculated.
The reference allocation code amount per block is calculated, and the data of the calculated reference allocation code amount per block of each color component is stored in the block allocation code amount data table. 1 for each color component in the block allocation code amount data table
The reference assigned code amount per block is given to the coding truncation circuit 16 when the corresponding color component is subjected to the variable length coding process.

The coding truncation circuit 16 adds the reference allocated code amount per block of the corresponding color component from the code amount allocating circuit 20 and the excess or deficiency of the generated code amount in the immediately preceding block to the allocated code amount. And the code length of each code from the code amount allocating circuit 20 is subtracted from the allocated code amount, and the remainder of the allocated code amount is calculated based on the total code amount of the code amount to be transmitted and the EOB code. If the value becomes smaller, a truncation signal is output and given to the variable length coding circuit 8, the coding of the block is completed, the transmitted code amount is subtracted from the allocated code amount of the block, and the value is assigned to the allocated code amount. It has the function of holding as the remainder.

Therefore, the encoding truncation circuit 16 refers to the allocated code amount, and if the input code amount to be transmitted and the code of the EOB are transmitted but the allocated code amount does not exceed the truncated value, the truncation is not performed, and the block is not terminated. The operation of terminating the encoding, subtracting the code amount to be transmitted from the allocated code amount of the block, and holding the value as the remainder of the allocated code amount is performed. Also, the encoding truncation circuit 16 refers to this allocated code amount, and if the allocated code amount is not enough to transmit the input DC component code to be transmitted and the EOB code, the two codes are determined. After the transmission, the coding of the block is terminated, and the operation of subtracting the transmitted code amount from the allocated code amount of the block, that is, maintaining the remainder of the allocated code amount as a negative value, is performed.

Reference numeral 10 denotes a code output circuit. The code output circuit 10 connects variable-length codes input from the variable-length coding circuit 8, and is configured by a recording medium such as a memory card. It functions to write to the recording system 22.

In this system, statistical processing is first performed using a standard quantization coefficient α for the initial time determined according to the shooting mode (first pass), and the amount of information for each color component necessary for optimization and the entire image Then, a process for performing optimized coding based on the information obtained by the statistical processing is started (second pass).

Therefore, first block the image, quantize the elements of the blocked image using a standard quantization coefficient α, variable-length encode the transform coefficients obtained by quantization, and Prediction of a coding coefficient α necessary for obtaining an optimum code amount from code amount information of each element of each color component obtained by long coding and code amount information of the entire image, a reference per block in each element of each color component Determining the amount of code to be allocated, shifting to a processing mode of optimal coding for a processing target image based on these, blocking the image in the execution of this processing mode, the predictive quantization coefficient for the elements of the blocked image A procedure such as quantization processing using α, variable-length coding of transform coefficients obtained by this quantization, and output processing for storing all codes of a processing target image is performed. Overall control managing shall we have to carry out by the control circuit 18 in FIG.
Note that such a function of the control circuit 18 can be easily realized by using a microprocessor (MPU).

 The above is the configuration of the encoding circuit 80.

The recording system 70 in FIG. 1 includes an interface circuit 70a and a recording medium 71 detachably connected to the interface circuit 70a. The image data encoded and output by the encoding circuit 80 and the quantization width (or corresponding to the Information) is recorded on the recording medium 71 via the interface circuit 70a.

Next, the operation of the present apparatus having the structure shown in FIGS. 1 and 2 will be described. First, the basic operation will be described with reference to FIG. 8, which is an operation transition diagram, in order to grasp the general outline.

When the camera user uses the camera, the user operates the switch 30 to set a desired number of recordable images. Thus, the control circuit 90 obtains the optimum code amount in accordance with the set number of recordable images and supplies the optimum code amount to the encoding circuit 80 as target code amount setting information. Thus, the number of shootable images is set.

Next, when photographing is performed, a subject image is formed as an optical image on an image sensor 40b located behind the photographing lens 40a.
Then, the imaging device 40b converts the formed optical image into an image signal and outputs the image signal. The image signal obtained by the image sensor 40b is input to the signal processing circuit 60, where the image signal is amplified by the amplifier circuit 60a in the signal processing circuit 60, A / D converted by the A / D converter 60b, and temporarily stored in the buffer memory 60c. Will be retained.
And after this, it is read from the buffer memory 60c,
Band correction by the process circuit 60d in the signal processing circuit 60,
Processing such as color signal formation is performed.

Here, since the subsequent encoding process is performed in the order of the Y (luminance), Cr, and Cb (all color difference) signals, the color signal formation is also performed accordingly. That is, the image signal is read out after being blocked in an 8 × 8 matrix, and the process circuit extracts a Y component, a Cr component (RY component), and a Cb component (BY component) from the blocked image signal data. ), The signals of these color components are separated, and gamma correction and white balance processing are performed.

The image signal data of each color component in the block image signal of the 8 × 8 matrix separated by the process circuit 60d is input to the encoding circuit 80. This gives 1
The image data for the frame (or for one field) is divided into blocks of the predetermined size and sequentially input to the encoding circuit 80. The image signal of each color component processed by the process circuit 60d may be stored in a buffer memory for each of Y, Cr, and Cb components, and may be read out and used in a later process.

In this embodiment, the Y component (luminance component) of the image signal data for one image is output from the signal processing circuit 60, and after this processing (statistical processing) at the subsequent stage is completed, the Cr component is then output. All the image data of the components are divided into blocks, and statistical processing is performed on the block at a later stage.
Thereafter, a process is performed in which the image of the Cb component is divided into blocks, and statistical processing in the subsequent stage is performed on the blocks.

The encoding circuit 80 supplies the input data received from the signal processing circuit 60 to the orthogonal transform circuit 4. Then, the orthogonal transform circuit 4 performs, for each block, two-dimensional orthogonal transform by, for example, discrete cosine transform (DCT) on the input image data (hereinafter, referred to as block image data) that has been divided into blocks. The orthogonal transform by the DCT is a process of dividing a certain waveform into frequency components and expressing this by the same number of cosine waves of different frequencies as the number of input samples.

The orthogonally transformed block image data (transformation coefficients) are stored at corresponding frequency component positions on an 8 × 8 matrix in a buffer memory (not shown). This is stored in a matrix having a relationship in which the frequency increases as the distance from the position increases), and this is input to the quantization circuit 6.

Then, the quantization circuit 6 performs the first pass (first pass) quantization on the block image data (transformation coefficient).
In the first quantization, an image quality set value set by a user upon photographing is applied to a predetermined quantization matrix for each frequency component (frequency components are determined corresponding to each matrix position of a block). , The transform coefficient is quantized with a quantization width obtained by multiplying the standard (temporary) quantization coefficient α given by the control circuit 18 (FIG. 8 (h1, i)).
The quantization matrix at this time may be the same for each of the luminance system and the chroma system. However, better results can be obtained by setting appropriate quantization matrices.

The quantized block image data (transform coefficients) is input to the variable length coding circuit 8, where it is subjected to variable length coding. The variable length coding circuit 8 performs a zigzag scan of the quantized and input transform coefficients in the order shown in FIG. 7 to scan from a low frequency component to a high frequency component. That is, the transform coefficients are stored in a 8.times.8 matrix corresponding to the frequency components. Since the frequency is lower as the position is closer to the origin, zigzag scanning enables scanning from lower frequency components to higher frequency components.

Since the first data in the scanning order in FIG. 7 is the DC component DC, the data of the DC component DC is the difference from the DC component DC of the block (the immediately preceding block) on which the variable-length coding was performed immediately before. The value diff-DC is Huffman-coded (see FIG. 8 (d
1), (e1)).

For the AC component AC, from the second scanning order in Fig. 7
Look at the transform coefficients in order up to the 64th, and when a transform coefficient that is not 0 (ie, valid) comes out, the number of consecutive 0 (invalid) coefficients that existed immediately before (zero run) and its validity Two-dimensional Huffman coding is performed with the coefficient values ((d2), (e2)).

Further, the variable-length encoding circuit 8 gives an EOB (end of block) code indicating the end of a block when an invalid coefficient continues from a certain coefficient to the 64th coefficient. Then, the code length generated for the code is output to the code amount calculation circuit 14 (g1). Then, such processing is executed for all blocks of one image.

After such processing for the Y component is completed, similar processing is performed for each of the Cr and Cb components.

On the other hand, the code amount calculation circuit 14 calculates the code amount of the entire image for each of the input Y, Cr, and Cb color components.
Calculation of the code amount for each color component of each of the Y, Cr, and Cb components and integration of the code amount for the entire image are performed (g2).
When all the Huffman encoding processes for one color component are completed, the data of the code amount for each color component is output to the code amount assignment circuit 20. The code amount assignment circuit 20 writes the code amount data for each color component as the code amount of the corresponding color component in the block assignment code amount table.

Then, at the stage where the Huffman encoding process for all the Y, Cr, and Cb components has been completed for all the blocks for one pixel, the code amount calculation circuit 14 controls the code amount data of the entire image under the control of the control circuit 18. The data is output to the quantization width prediction circuit 12 and the data of the code amount of the entire image is
Output to

From the input code amount data of the entire image and the target total code amount data, for example, the quantization width prediction circuit 12 determines the optimum quantization coefficient α to approximate the target total code amount by linear prediction. (H2) in FIG. 8 based on the quantized coefficients used in (1).

Further, the code amount allocating circuit 20 determines a reference allocated code amount per block of each color component from the input code amount of each color component, the code amount of the entire image, and the target total code amount, for example, the code amount of each color component. The target total code amount is calculated, for example, by proportional distribution using the ratio of the amounts (FIG. 8 (h3)).

Specifically, to determine the reference assigned code amount per block of a certain color component, the code amount of the color component is multiplied by the target total code amount, and the result is divided by the code amount of the entire image. The obtained result is divided by the number of blocks in the color component to obtain a reference assigned code amount per block of the color component. Then, the calculated data of the reference assigned code amount per block of each color component is stored in the block assigned code amount data table. The data of the allocated code amount for each color component in the block allocated code amount data table is supplied to the coding abort circuit 16 when the corresponding color component is subjected to the variable length coding process.

Thus, the first pass, that is, the first encoding (statistical processing) for determining the code amount to be allocated to each block and optimizing the quantization width is completed.

Next, the process enters the second pass. The process of the second pass is the second encoding (encoding process), and is a process of obtaining a final encoded output optimized to be within the target total code amount.

This process is first performed for the Y component, and is performed for the Cr and Cb components after the Y component is completed. That is, first, the image data is divided into blocks and read out, and the Y component (luminance) image signal data extracted and output from the signal processing circuit 60 is input to the encoding circuit 80 (a). The input block image data is encoded by an encoding circuit.
The signal is input to the orthogonal transformation circuit 4 in 80 and the orthogonal transformation is performed again (b). The transform coefficient obtained by this orthogonal transform is input to the quantization circuit 6, where the quantization is performed again (c). However, the quantization coefficient α used at this time is the optimum quantization coefficient α for the prediction calculated by the quantization width prediction circuit 12 in the previous pass.

Next, the transform coefficient of the quantized block image data is
It is input to the variable length coding circuit 8. In the variable length coding, as in the case of the statistical processing, of the conversion coefficients of the block image data, first, the difference value diff-DC of the DC component DC is Huffman-coded ((d1), (e1)), and then the AC component The AC is sequentially extracted by zigzag scanning to perform two-dimensional Huffman coding ((d2), (e2)).

However, every time a Huffman code for one element (one position in the matrix) is generated, the code amount allocating circuit 20 outputs one block per one block to be transmitted in the element stored in the block allocated code amount data table. The reference allocated code amount is output to the coding truncation circuit 16. On the other hand, the coding truncation circuit 16 adds the reference allocated code amount per block of each color component and the remainder to the allocated code amount in the coding process in the previous block. If the allocated code amount is not exceeded even if the code amount to be transmitted and the code of the EOB are transmitted, the truncation signal is not generated and the block is transmitted from the allocated code amount. A process for reducing the power amount is performed.

Then, when the total code amount of the code amount of the block to be transmitted and the code of the EOB exceeds the remaining code amount of the allocated code amount, the coding truncation circuit 16 truncates to the variable length coding circuit 8. A signal is output, Huffman coding of the block is terminated, and the transmitted code amount is subtracted from the allocated code amount of the block, and the value is retained as the remainder of the allocated code amount. Then, the variable length encoding circuit 8 shifts to the Huffman encoding of the next block obtained by the quantization circuit 6.

Further, in the coding truncation circuit 16, when the allocated code amount is not enough to transmit the DC component and the code of EOB, the code truncation circuit 16 transmits the DC amount and the EOB code, and then the coding truncation circuit 16 sets the variable length coding circuit. 8 to terminate the Huffman encoding of the block, subtract the transmitted code amount from the allocated code amount of the block, and retain the remainder of the allocated code amount as a negative value.

Therefore, the variable length encoding circuit 8 outputs the converted Huffman code to the code output circuit 10 until the truncation signal is input from the encoding truncation circuit 16, and outputs the Huffman codes for all the elements of the matrix before the truncation signal is generated. When the encoding is completed, the variable length encoding circuit 8 outputs the code of the EOB to the code output circuit 10. Also, if the truncation signal is input before the Huffman coding for all elements of the matrix is completed, the variable-length coding circuit 8 outputs an EOB code to the code output circuit 10 instead of the code. Become. The code output circuit 10 temporarily stores the coded data.

Then, the variable length encoding circuit 8 shifts to the Huffman encoding of the next block obtained by the quantization circuit 6.

Such an operation is repeated, and the processing of all the blocks of the image of one screen is completed, thereby completing all the encoding processing. When such processing for the Y component is completed, processing for chroma-based components (Cr, Cb) is then started in the same manner. Even in the processing of the chroma component, the quantization circuit 6 calculates the optimum quantization coefficient α of the prediction calculated by the quantization width prediction circuit 12 in the previous pass.
Use

When the processing of the second pass is completed for all the blocks of the image for one screen for the chroma components, all the encoding processes are completed.

At the end of this, the code output circuit 10 outputs the optimized Huffman coded data for one image to the recording system 22,
The data is written to a storage medium 71 such as a memory card in the recording system 22 (f). This is performed by the output of the code output circuit 10. The code output circuit 10 connects the Huffman codes, which are the variable length codes from the variable length coding circuit 8, and writes them by giving them to a memory card as a storage medium 71. . The writing to the storage medium 71 by the output of the code output circuit 10 may be performed collectively at the end of the second pass, but may be changed at the end of the first pass and the start of the second pass. The results obtained by joining the long Huffman codes may be written to the storage medium sequentially in units of one or several bytes, as soon as they are collected.

Prior to this, the code output circuit 10 writes the optimum quantization coefficient α used for coding in the header portion of the storage data of the coded image, and leaves it as a clue at the time of reproduction.

As described above, in the present apparatus, the reference allocated code amount per block is determined in accordance with the total code amount per target image (target total code amount) determined according to the number of shootable images. The amount of generated code for the block is adjusted by adjusting the excess or deficiency of the generated code amount with respect to the allocated code amount of the block, and the generated code amount of the entire image is controlled by using this to control the generated code amount of each block. Is designed not to exceed the target total code amount, which is an important point of the present invention.

Therefore, the present invention is not limited to the block size, the type of orthogonal transform, the type of variable length coding, and the like used in the present embodiment. Further, the statistical processing need not always be performed once, but may be omitted, or may be performed a plurality of times.

However, by optimizing the quantization width by performing statistical processing, the effect of the present invention becomes particularly remarkable. In addition, considering the processing time and the obtained effect, it is necessary to perform statistical processing once. Is the best solution. In statistical processing,
Instead of obtaining the total code amount for each color component and using this to determine the code amount per block, the code amount may be empirically weighted for each color component. Without weighting, the code amount per block may be the same, and the statistic amount in the statistical processing is not always necessary for determining the reference allocated code amount per block. Further, the statistical processing is not limited to the provisional quantization width, and is not limited to the processing of actually performing the encoding to obtain the code amount, but may be the calculation by calculating the activity, for example.

Also, the allocated code amount per block in each color component does not necessarily need to be uniform, but is weighted by the block position with respect to the image. For example, a larger code amount is assigned to a block at the center of an image than to peripheral blocks. You may make it assign. Further, as shown in the present embodiment, it is not always necessary to allocate the target total code amount to the reference allocation code amount per block. For example, the target code amount is allocated to the reference allocation code amount per block. The remainder that is not divisible when the first block may be given to the first block as an excess or deficiency up to the previous block.

Also, the compression rate may not be variable, and may be fixed to one type of compression rate, and the target total code amount, the quantization width, the reference allocation code amount per block, and the like may all be given as fixed values. In this case, the configuration becomes simpler. Further, the image data buffer memory may be provided between the orthogonal transformation circuit 4 and the quantization circuit 6, and in this case, the processes of blocking and orthogonal transformation in the encoding process can be omitted. However, in order to maintain accuracy, the size of the image memory is increased in this case.

Also, process processing should be performed before A / D conversion,
After that, it may be digitized. In addition, in the present apparatus, variable-length coding for each block is performed from low-frequency components, and high-frequency components having relatively little visual effect on image quality are omitted by terminating the coding. It is possible to perform encoding with high compression while minimizing deterioration.

The present invention having the configurations shown in FIGS. 1 and 2 described above is, in short, a basic allocation code amount per block determined in advance from the target total code amount and a code amount generated in a block before the block. When the code amount is monitored by monitoring the code amount for each block so that the code amount falls within the allocated code amount obtained by adding the excess or deficiency to the allocated code amount, when the code amount reaches the allocated code amount, Although the coding of the block is terminated, the DC component is always coded even if it exceeds the allocated code amount, the excess or deficiency of the allocated code amount at that time is stored, and the process proceeds to the coding of the next block. Things. This is because the amount of generated code in each block of the image is different, that is, although there is a local shift in the amount of code,
A block of several blocks or tens of blocks is a micro block (several blocks of an 8 × 8 matrix)
It is noted that the generated code amount for each microblock has a relatively small deviation.

Therefore, instead of adaptively allocating the code amount for each block, the method of allocating the allocated code amount per block is fixed, and the target code amount is allocated according to the allocation method, and the reference allocated code per block is allocated. Even if the amount is determined, the surplus amount of the generated code amount with respect to the allocated code amount is carried over to the allocated code amount of the subsequent blocks, thereby adapting to the change in the generated code amount of each block. You are controlling the amount.

In order to adaptively allocate the allocated code amount for each block to each image, an operation of obtaining a statistic for each block for each image and the like, and storing a reference allocated code amount for each block obtained as a result, In this method, these are not required, and even if the compression ratio changes,
The same hardware can cope with it, and therefore, it is not necessary to prepare hardware for each target compression rate (for each code amount), so that it is possible to reduce the cost and size of the applied device. Furthermore, the allocated code amount of the block is determined by adding the excess or deficiency in the previous block to the reference allocated code amount, and the coding is controlled based on this, so that the processing can be simplified, and therefore the processing time can be shortened. it can.

The above-described encoding circuit 80 in the configuration of FIG. 2 performs a series of processes in the first pass with the calculated tentative quantization width based on the target code amount in the compression encoding. The second pass is performed using the optimum quantization width in order to obtain the final compression-encoded data. This is to find the optimum α.

In FIG. 3, the signal flow in the first pass is indicated by a dotted arrow, and the signal flow in the second pass is indicated by a solid arrow in order to make the processing flow of the encoding circuit 80 easy to understand. Is shown. The following is a brief description of the operation following the flow of this signal.

When the image data is encoded, the target total code amount is set in the control circuit 18 of the encoding circuit 80. This is because the user sets the desired number of recordable images by operating the switch 30, and the control circuit 90 selects the optimal code amount according to the set number of recordable images, and uses this information as the target total code amount information. Is provided to the encoding circuit 80. In the initial state, the number is set to a predetermined standard photographable number.

When photographing is performed, an image signal is output from the image sensor in the image pickup system 40. This output image signal is converted into a digital signal in the signal processing circuit 60,
After being stored in the buffer memory, it is read out in blocks of 8 × 8 pixels and separated into a Y component, then a Cr component, and then a Cb component. This separation is first performed on the Y component,
The image data of the Y component output in units of 8 × 8 pixels is input to the orthogonal transform circuit 4 and orthogonally transformed (DCT; Discrete Cosin in this example) for each block.
e Transform) is performed.

The DCT transform coefficients obtained by the orthogonal transform circuit 4 are input to the quantization circuit 6, while the control circuit 18 outputs the target total code amount to the quantization width prediction circuit 12, and the quantization width prediction circuit 12 From the code amount, the quantization coefficient α is calculated using the relationship of the equation (1).
Is set and output to the quantization circuit 6. The quantization circuit 6 linearly quantizes the transform coefficient using the input quantization coefficient α. The quantized transform coefficients are input to the variable-length coding circuit 8, where variable-length coding (Huffman coding in this example) is performed.

The input quantization coefficient is scanned from a low-frequency component to a high-frequency component called zigzag scan, and the first DC component data is the same as the DC component of the block that has just been subjected to variable length coding. Are subjected to Huffman coding and output.

For the AC component, the conversion coefficients are sequentially examined from the second to the 64th in the scanning order. If a conversion coefficient that is not 0 (that is, a valid coefficient) appears, a continuous 0 (zero; Two-dimensional Huffman coding is performed with the value of (invalid) coefficient (zero run) and its effective coefficient.
If the invalid output continues to the 64th coefficient after a certain coefficient, the code of the EOB indicating the end of the block is output. The variable length coding circuit 8 performs the above-described coding, and outputs the code length to the code amount calculation circuit 14 every time one code is generated.

The code amount calculation circuit accumulates and stores the input code lengths for each color component.

When such processing for the Y component is completed,
The same processing is performed for the Cr component and the Cb component.

When encoding of one image is completed, the code amount calculation circuit 14 adds the accumulated code amounts of the respective color components and calculates the code amount of the entire image as a total code amount value. The total code amount value is output to the quantization width prediction circuit 12, and is output to the code amount allocation circuit 20 for the code amount of each color component and the entire image.

When the above-described first pass encoding process is completed, the quantization width prediction circuit 12 calculates the total image code amount obtained by the encoding in the first pass and the target code amount given from the control circuit 18, A more suitable quantization coefficient α is predicted, and the quantization circuit 6
Output to The code amount assignment circuit 20 is a code amount calculation circuit.
From the code amount of each color component, the code amount of the entire image, the target code amount, and the number of blocks, which are input from step 14, the allocated code amount per block for each color component is calculated.

Subsequently, a second pass encoding process is performed on the same image data. In the second pass, the image data read from the memory in the signal processing circuit 60 is first separated into a Y component, then a Cr component, and then a Cb component, and the image data of each component is 8 × 8 pixels. After processing such as blocking is input to the orthogonal transform circuit 4, the orthogonal transform (DCT) is performed for each block.
Conversion), whereby the DC obtained by the orthogonal transformation circuit 4 is obtained.
The T transform coefficients are input to the quantization circuit 6.

In the quantization circuit 6, the corrected quantization width by the new quantization coefficient α based on the given prediction is used to calculate
Linearly quantizes the DCT transform coefficients. The quantized coefficients are input to the variable length coding circuit 8, and are subjected to Huffman coding in the same manner as in the first pass coding. Here, the amount of code generated at the time of encoding is obtained at the end of the first pass, and the amount of code per block stored in the code amount allocating circuit 20 and the amount of code allocated in the previous block are determined. The comparison with the added code amount of the block is performed, and if the amount exceeds the value, the subsequent code is stopped in the block by the operation of the code cutoff circuit 16. Further, the excess or deficiency of the allocated code amount at the time when the encoding of the block is completed is stored in the code amount allocating circuit 20.

The coded data controlled to the target total code amount by the above method is sequentially output to the recording system 70 via the code output circuit 10 and recorded.

Next, reproduction of the compression-encoded recording image data of the recording medium 71 recorded by the recording system 70 will be described.

FIG. 4 shows the configuration of the regenerator. In the figure, reference numeral 100 denotes a playback device main body, which includes a reading unit 102, a decoding circuit 104, a processing circuit 106, and a control circuit 108. The reading unit 102 can attach and detach the storage medium 71, and reads the contents of the storage medium 71 via the interface circuit 110.

The decoding circuit 104 has functional blocks as shown in FIG. That is, 112 is a Huffman decoding unit that decodes the Huffman coded data, and 114 is the Huffman-decoded data read out from the storage medium 71 based on the quantization width information read out from the storage medium 71 and set and input. An inverse quantization unit for quantizing 116, an IDCT (inverse DCT transform) unit for inverse DCT transforming the data obtained by the inverse quantization and outputting it as video signal data, and a control unit 118 for controlling these Department.

The processing circuit 106 includes a buffer memory 120, an encoder 122, and a D / A converter 124. The buffer memory 120 is a memory for temporarily holding the video signal data output from the decoding circuit 104, and the encoder 122
The D / A converter 124 converts the video signal data read from the video signal 20 into an NTSC video signal, and converts the NTSC video signal into an analog signal and outputs it as a TV video signal. is there.

The control circuit 108 controls the entire reproduction apparatus main body 100, and controls the reading unit 102 of the reproduction apparatus main body 100 to read the information of the quantization width at the time of encoding.
The information of the quantization width at the time of encoding read from the recording medium 71 is set in the inverse quantization unit 114 of the decoding circuit 104, and then the control circuit 108 controls the video signal data compressed and encoded from the recording medium 71. Is controlled to control the reading unit 102 in order to read the data. Although not shown, the reproducing apparatus main body 100 has a frame feed switch and the like, and can reproduce a video at a frame position designated by the switch. The control circuit 108 also performs such control.

Next, the operation of the reproducing apparatus having the above configuration will be described. When the recording medium (memory card) 71 on which the compression-encoded video signal data is recorded is mounted on the reading unit 102 of the player main body 100,
First, the control circuit 108 controls the reading unit 102 to read information on the quantization width at the time of encoding. As a result, the reading unit
In 102, information on the quantization width at the time of encoding is read from the recording medium 71, and this information is used as the inverse quantization unit 1 of the decoding circuit 104.
Set to 14. Subsequently, since the control circuit 108 controls the reading unit 102 to read the video signal from the recording medium 71, the reading unit 102 sequentially reads the video signal from the recording medium 71 and inputs the video signal to the decoding circuit 104. In the decoding circuit 104 receiving this, the Huffman decoding unit 112 decodes the Huffman code to obtain a quantized coefficient. The quantization coefficient thus obtained is supplied to the inverse quantization circuit 114 to be inversely quantized. Here, the inverse quantization is performed by using the information of the quantization width previously set.

The transform coefficient obtained by the inverse quantization is subjected to inverse DCT for each block in the IDCT unit 116, and restored to the original video signal. In this way, the video signal is restored in the order of Y, Cr, and Cb, output from the decoding circuit 104, and written to the buffer memory 120 in the processing circuit 106. When the writing of the video signal data for one screen is completed, the video signal data is read out from the buffer memory 120 in the normal scanning order of the television signal, and is converted into an NTSC video signal by the encoder 122. Further, the signal is converted into an analog signal by the D / A converter 124 and output. By inputting this video signal to a television monitor, the image is reproduced as a television video and can be viewed as a video, and a hard copy can be obtained by giving it to a printing device such as a video printer and printing it. It can be viewed in the same way as.

As described above, the camera can set a desired number of shootable images, and by setting the number of shootable images, the camera automatically sets a compression rate corresponding to the settable number of shots and temporarily sets a compression rate corresponding to the set compression rate. The quantized image data for one screen is quantized using a variable quantization width, and is subjected to variable length encoding. The optimal quantization width is predicted from the code amount of the captured image data for one screen obtained as a result. Quantizing the captured image data for one screen by the optimized quantization width and performing variable-length encoding, and decoding the encoded video signal using the optimal quantization width used at the time of shooting when reproducing the encoded video signal. Accordingly, encoding at a desired compression rate can be realized by one common hardware without providing hardware for each compression rate, and similarly, decoding of image data encoded at a desired compression rate can be performed by compression. Without providing any hardware rate by, it can be realized by a single common hardware.

In the embodiment, a two-pass method is used in which the encoding process is completed in two processes of a first pass and a second pass. However, the present invention is not limited to this, and a quantization width set from a compression ratio is used. 1
In practice, a sufficient image can be obtained even if the coding is performed in one pass, and a better result can be obtained even if the first pass is repeated several times.

In addition, although an example in which a memory card is used as a recording medium has been described, a floppy disk, an optical disk, a magnetic tape, or the like may be used. Although the camera and the playback device are shown as being separated from each other, the camera may be of an integrated type having the functions of the playback device. Furthermore, in the embodiment, the quantization width value itself is recorded on the recording medium. However, the quantization width value may be converted or encoded and recorded.

In the above embodiment, the quantization width is set based on the target total code amount. However, in an application in which a plurality of target total code amounts are switched and used in a mode, the quantization width and the 1st corresponding to each mode are used. Of course, it is of course possible to prepare an allocated code amount per block in advance and use it by switching between modes.

The present invention is not limited to the embodiments described above and shown in the drawings, and can be appropriately modified and implemented without departing from the scope of the present invention. , Can be applied to compression encoding of various images.

Further, in the above embodiment, information of the total code amount is given from the compression ratio corresponding information, and a quantization coefficient capable of giving a quantization width corresponding to the total code amount is predicted, and the predicted quantization coefficient is Although the quantization is performed with the quantization width based on, the relation of the quantization width with respect to the compression ratio correspondence information is calculated in advance and tabulated, and this is stored in a memory or the like, and the compression ratio correspondence information is stored. It is also possible to directly output quantization width information (that is, quantization coefficient or quantization width value). In this way, when the information corresponding to the desired compression ratio is input, the information of the quantization width uniquely determined corresponding to the input compression ratio corresponding information can be read and output, and the optimum quantization can be immediately performed. As soon as the width can be set, the quantization circuit can quantize the image signal data with this quantization width. The same applies to the reference assigned code amount per block.

According to this configuration, the information of the optimal quantization width corresponding to various compression ratios and the reference allocation code amount per block are stored in a memory or the like in advance, and are read out corresponding to the input compression ratio corresponding information. Alone can provide information on the optimal quantization width and the reference allocation code amount per block, so that when encoding to fit into the target total code amount, the processing can be performed in a very short time, and hardware It will be easy.

〔The invention's effect〕

As described in detail above, according to the present invention, it is possible to control the code amount and to perform the encoding process in a short time without requiring large-scale hardware. Therefore, it is possible to provide an image data encoding apparatus and an encoding method that are optimal for an electronic camera system and the like.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of the present invention, and FIG.
FIG. 3 is a block diagram showing an example of the configuration of a main part of the apparatus of the present invention, FIG. 3 is a block diagram for explaining the operation flow of the circuit of FIG. 2, and FIGS. FIG. 6, FIG. 6 is a perspective view showing the appearance of the electronic camera body according to the present invention, FIG. 7 is a diagram for explaining zigzag scanning of a block divided into 8 × 8 pixels, and FIG. FIG. 9 is an operation transition diagram for explaining the principle operation, and FIG. 9 is an operation transition diagram for explaining the prior art. DESCRIPTION OF SYMBOLS 1 ... Electronic camera main body, 6 ... Quantization circuit, 8 ... Variable length encoding circuit, 10 ... Code output circuit, 12 ... Quantization width prediction circuit, 14 ... Code amount calculation circuit, 16 ... Encoding truncation circuit, 18, 18a ... control circuit, 20 ... code amount allocation circuit, 24 ...
... DCPM circuit, 30 ... Switch, 40 ... Imaging system, 48 ... LC
D display, 60: Signal processing circuit, 80: Encoding circuit, 70
... recording system, 71 ... recording medium, 90 ... control circuit, 100 ...
... Reproducer body, 102 ... Reading unit, 104 ... Decoding circuit, 10
6 Processing circuit, 108 Control circuit, 110 Interface circuit, 112 Huffman decoding unit, 114 Inverse quantization unit, 116 IDCT (inverse DCT transform) unit, 118 Control unit.

Continuation of the front page (56) References JP-A-2-104078 (JP, A) Chikako Mogi, et al., "Code amount control method in DCT coding", Proc. Of the 1990 IEICE Spring Conference. , The Institute of Electronics, Information and Communication Engineers, March 1990, Volume 7, p. 62 (D-310) (58) Field surveyed (Int. Cl. ⁶ , DB name) H04N 1/41-1/419 H04N 7/24-7/68

Claims (6)

(57) [Claims] An image data is divided into blocks, each of the divided blocks is subjected to an orthogonal transform in turn, and the transform output is quantized by a quantizing means, and the quantized output is supplied to a variable length coding means. An encoding apparatus that performs variable-length encoding on each block and encodes data for each block. In encoding the block, the amount of code generated by the encoding process in the block to be encoded and the allocation in the block. Calculate the excess or deficiency of the generated code amount with respect to the allocated code amount based on the code amount and divide a predetermined code amount equal to or less than a target total code amount which is an allowable code amount per image by the total block number. Thus, the code amount obtained by adding the calculated excess / deficiency amount to the predetermined reference code amount per block is determined as the allocated code amount of the next block to be processed. Allocating means; and truncating means for controlling the variable-length coding means to terminate the variable-length coding so as not to exceed the allocated code amount allocated by the code amount allocating means for each block. An image data encoding device characterized by the above-mentioned. 2. The image data is divided into blocks, and each of the divided blocks is subjected to an orthogonal transform in order to decompose the image data for each block from low frequency components to high frequency components in order. The output is quantized by a quantizing unit, and the quantized output is supplied to a variable-length encoding unit to perform variable-length coding, and encodes data for each block. Based on the code amount generated by the encoding process in the block to be encoded and the allocated code amount in the block, an excess or deficiency amount of the generated code amount with respect to the allocated code amount is calculated, and an allowable amount per image is calculated. By dividing a predetermined code amount equal to or less than the target total code amount, which is the code amount, by the total number of blocks, the calculated reference code amount per block is calculated as described above. Code amount allocating means for obtaining the code amount obtained by adding the excess or deficiency amount as the allocated code amount of a block to be processed next, and the code amount allocated by the code amount allocating unit for each block so as not to exceed the allocated code amount. Aborting means for controlling the variable-length encoding means to abort the variable-length coding, and a control means for controlling the quantization means to perform quantization using a predetermined quantization width, The variable-length encoding means obtains a difference between the DC component of the image data of each block obtained by the orthogonal transformation and the DC component of the preceding encoding processing block, and sets the difference to a variable length. An encoding processing unit for a DC component to be encoded and an encoding processing unit for an AC component for performing variable-length encoding on the AC component after the processing of the encoding processing unit for the DC component are completed. The coding truncation means coding device of the image data, characterized by a configuration allowing only abort control the encoding processing section for the AC components. 3. The image data is divided into blocks, and each of the divided blocks is subjected to orthogonal transform in order to decompose the image data of each block from low frequency components to high frequency components in order. In an encoding device, the output is quantized by a quantizing unit, the quantized output is supplied to a variable-length encoding unit and subjected to variable-length encoding, and data is encoded for each block. Code amount calculating means for calculating the code amount of the transform coefficient of the entire transformed image, and the optimal quantum based on the code amount of the entire image calculated by the code amount calculating means and the permissible target total code amount per screen. A quantization width prediction unit that predicts a quantization width and gives the prediction quantization width to the quantization unit; and, in encoding the block, codes in a block to be encoded. Based on the amount of code generated by the processing and the amount of code allocated in the block, the amount of excess or deficiency of the amount of generated code with respect to the allocated amount of code is calculated, and a predetermined amount of code equal to or less than the target total amount of code is calculated as a total Code amount allocating means for obtaining a code amount obtained by adding the calculated excess / deficiency amount to the determined reference code amount per one block as the allocated code amount of a block to be processed next by dividing by a number; The truncation means for controlling the variable length coding means so as to terminate the variable length coding so as not to exceed the code amount allocated by the code amount allocation means; A first encoding process of estimating an optimal quantization width from a total code amount using a predetermined temporary quantization width is performed at least once, and then the quantization width prediction unit performs And a control means for performing control so as to execute the second coding process using the predicted optimum quantization width. 4. The image data is divided into blocks, and each of the divided blocks is subjected to an orthogonal transform in order to decompose the image data for each block from low frequency components to high frequency components in order. In the encoding device, the transform output is quantized by a quantizing unit, and the quantized output is supplied to a variable length encoding unit to perform variable length encoding, and encodes data for each block. Code amount calculating means for calculating the code amount of the transform coefficient of the entire coded image; optimal code amount based on the code amount of the entire image calculated by the code amount calculating means and the permissible target total code amount per screen. A quantization width prediction unit that predicts a quantization width and provides the prediction quantization width to the quantization unit; and, in encoding the block, a code in a block to be encoded. Calculating the excess or deficiency of the generated code amount with respect to the allocated code amount based on the code amount generated by the conversion process and the allocated code amount in the block, and calculating a predetermined code amount equal to or less than the target total code amount. Code amount allocating means for calculating a code amount obtained by adding the calculated excess / deficiency amount to the determined reference code amount per block by dividing by the number of blocks as the allocated code amount of the next block to be processed; The truncation means for controlling the variable length coding means so as to terminate the variable length coding so as not to exceed the code amount allocated by the code amount allocation means for each block; Performs a first encoding process of predicting an optimum quantization width from the target total code amount using a predetermined temporary quantization width at least once, and then executes the quantization width prediction means. And control means for performing the second encoding process using the optimal quantization width predicted by the above, wherein the variable-length encoding means includes a unit for each block obtained by the orthogonal transform. Among the image data, for the DC component, an encoding processing unit for a DC component that obtains a difference from the DC component in the preceding encoding processing block and performs variable-length encoding of the difference, and an encoding process of the DC component After the processing of the unit is completed, an AC component encoding processing unit that performs variable-length encoding on the AC component is provided, and the truncation unit enables the discontinuation control only for the AC component encoding processing unit. A coding device for image data, wherein 5. The image data is divided into blocks, each of the divided blocks is subjected to orthogonal transform in turn, and the transform output is quantized by a quantizing means, and the quantized output is supplied to a variable length coding means. An encoding apparatus that performs variable-length encoding on each block and encodes data for each block. In encoding the block, the amount of code generated by the encoding process in the block to be encoded and the allocation in the block. The amount of excess or deficiency of the generated code amount with respect to the allocated code amount is calculated based on the code amount, and the code amount for each color of the target total code amount, which is the allowable code amount per image, is calculated for the color. By dividing by the number of blocks per minute,
Code amount allocating means for obtaining a code amount obtained by adding the calculated excess / deficiency amount to a predetermined reference code amount per block as the allocated code amount of a block to be processed next; And censoring means for controlling the variable-length coding means so as to stop the variable-length coding so as not to exceed the allocated code amount allocated by (1). 6. The apparatus according to claim 1, wherein said amount allocating means allocates a larger amount of code to a block in a central portion of the image than a peripheral block.
An image data encoding apparatus according to any one of the preceding claims.