JP3249616B2 - Image data compression apparatus and method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to, for example , lossless compression of an image.
The present invention relates to an image data compression apparatus and method including a plurality of compression means including a reversible compression means for selecting an optimum compression means according to input image data.

[0002]

2. Description of the Related Art In recent years, images created on a computer by DTP or the like have been required to have higher image quality. Has become. For this reason, computer-generated images are treated as code information in a page description language or the like to reduce the amount of information. However, there is a problem that it takes time to develop the code information into the image data and that the original image data cannot be reproduced even if the code information is developed.
Therefore, image data is reduced by performing compression using a reversible compression method that maintains the uniformity of colors and the sharpness of outlines of a computer-generated image.

[0003]

However, in the above conventional example, even if an image (hereinafter, a natural image) or a gradation image read by a scanner and inserted into an image created by a computer is processed by the above-described reversible compression method, The compression ratio did not increase much, and in some cases increased slightly. Natural images and gradation images are low-frequency components compared to ordinary computer-generated images that have important high-frequency components, such as outline clarity and color uniformity. The reversibility is not so important, and there is a characteristic that even if the data changes due to the irreversible compression / decompression processing, there is not much deterioration.

[0004] Therefore, it is necessary to change the compression method in a portion where the compression ratio does not increase so much by the reversible compression method, and where the deterioration is not conspicuous even when the compression is performed by the irreversible compression method, and other portions. The present invention has been made to solve the above problems, pressure
An object of the present invention is to select a desired compression method in accordance with the characteristics of each block to be reduced, thereby improving the compression efficiency of the block and performing high-quality compression. Special
Then, as a lossless compression method, the block to be compressed is
It is an algorithm that compresses every pixel
Specifically, the compression target pixel is
Compress based on whether it matches the pixel data
, Each block can be efficiently compressed
The purpose is to be. Furthermore, the above-mentioned reversible compression method and
In controlling the selection of the irreversible compression method,
Is each block one-dimensionally arranged horizontally or vertically?
Select with high accuracy, taking into account that compression
The purpose is to be able to do.

[0005]

To achieve the above object of the Invention The present invention is an image data compression apparatus for compressing image <br/> image data for each block comprising a plurality of pixels data, the
A pixel data in which a plurality of pixel data in a block are arranged one-dimensionally
The data sequence is sequentially set as compression target pixel data, and the compression target pixel data is
Compare data with previously compressed pixel data
Reversible compression means for performing reversible compression based on the result;
Image data using an algorithm different from the compression means
Irreversible compression means for irreversibly compressing each block,
A pixel data in which a plurality of pixel data in a block are arranged one-dimensionally
For a data row, different pixel data for each predetermined number of pixel data
Data is detected, and before each block is
The respective lossless compression means based on the detection result.
Or any one of the irreversible compression means.
And control means for controlling . Also book
According to the invention, the image data is divided into blocks each including a plurality of pixel data.
A method of compressing image data, comprising:
Pixel data sequence in which multiple pixel data in a block are arranged one-dimensionally
Are sequentially set as compression target pixel data, and the compression target pixel data
And the result of comparing the compressed pixel data with
A lossless compression step of performing lossless compression based on the
The image data using an algorithm different from
An irreversible compression step for irreversible compression for each block;
Pixel data sequence in which multiple pixel data in a block are arranged one-dimensionally
Of pixel data that differs for each predetermined number of pixel data
Detecting the number of existing blocks and detecting the number corresponding to each block
Based on the result, each block is subjected to the lossless compression process or the
Control to compress in any of the lossy compression processes
And a control step.

[0006]

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A preferred embodiment according to the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram illustrating a rough functional configuration of the data compression device according to the embodiment. In the figure, reference numeral 1 denotes a pixel data holding unit, which holds 24 bits of pixel data per pixel for 8 × 8 pixels in the horizontal and vertical directions. A switch 2 distributes the input 8 × 8 pixel data in accordance with a control signal described later. Reference numeral 3 denotes an area determination unit that determines an area from data input from the pixel data holding unit 1 and outputs a control signal. Reference numeral 4 denotes a reversible compression unit which converts input data into compressed data and outputs the data. Reference numeral 5 denotes an ADCT compression processing unit that performs ADCT compression processing on input data and outputs the data. Reference numeral 6 denotes a memory unit that stores the compressed data converted by the reversible compression unit 4 and the ADCT compression processing unit 5 and the above-described control signal at an address specified by the control unit 7 described later. 7
Denotes a control unit, which controls the operation timing of each component described above. In the embodiment, the pixel data holding unit 1 holds 8 × 8 pixels in the horizontal and vertical directions, but the present invention is not limited to this.

FIG. 2 is a block diagram showing a detailed configuration of the area determining section 3 shown in FIG. In the figure, reference numeral 201 denotes a parallel / serial converter, which inputs parallel pixel data output from the pixel data holding unit 1 in units of eight pixels, converts the data into serial data, and outputs the serial data. Reference numeral 202 denotes a data processing circuit which outputs "1" when four or more different color data out of eight pieces of input pixel data exist. Reference numeral 203 denotes an adder, and the eight data processing circuits 2
The output value from 02 is added. 204 is a comparator,
The output value from the adder 203 is compared with a predetermined value (“4” in the embodiment), and if it is less than the predetermined value, “0” is output, and if it is more than the predetermined value, “1” is output.

FIG. 3 shows a data processing circuit 202 shown in FIG.
FIG. 3 is a block diagram showing a detailed configuration of one of FIG. In the figure, reference numeral 301 denotes a latch, which sequentially stores input pixel data. Latches 302 to 304 sequentially store pixel data in synchronization with a control signal 3a described later. 3
Reference numerals 05 to 307 denote comparators, which compare the pixel data stored in the latch 301 with the pixel data stored in the latches 302 to 304, respectively, and set “0” when they are the same and “1” when they are different. Output. Reference numerals 308 and 309 denote selectors, which selectively output the outputs from the comparators 306 and 307 or “1” according to control signals 3 b and 3 c described later. Reference numeral 310 denotes a counter, and the output from the comparator 305 and the outputs from the selectors 308 and 309 are all “1”.
The control signal 3a output in the case of is counted and output.
Reference numeral 311 denotes a comparator which compares the output from the counter 310 with a predetermined value (“4” in the embodiment), and outputs “0” when the value is less than the predetermined value and outputs “1” when the value is more than the predetermined value. I do. A decoder 312 decodes an output from the counter 310 and outputs control signals 3b and 3c for the selectors 308 and 309.

FIG. 4 is a block diagram showing a detailed configuration of the lossless compression section 4 shown in FIG. In the figure, reference numeral 401 denotes a pixel data rearranging unit, which is a pixel data holding unit 1 shown in FIG.
And serially rearranges the parallel pixel data input from. A latch 402 stores the output from the pixel data rearranging unit 401 with a delay of one pixel. A latch 403 stores an output from the latch 402 according to the signal 4a. Reference numeral 404 denotes a comparator which compares the pixel data output from the pixel data rearranging unit 401 with the pixel data from the latch 402, and outputs "0" if they are the same and "1" if they are different. 4
A comparator 05 compares the pixel data from the pixel data rearranging unit 401 with the pixel data stored in the latch unit 403 in the same manner as the comparator 404.
Are output, and "1" is output when they are different. Reference numeral 406 denotes a selector, which outputs the signal 4 from each of the comparators 404 and 405.
a, 4b and the output from the pixel data rearranging unit 401 are input, and data is selected and output under the control of the control unit 407. Reference numeral 407 denotes a control unit which controls the signals 4a, 4
b is input to control the selector 406.

FIG. 5 is a block diagram showing a detailed configuration of ADCT unit 5 shown in FIG. In the drawing, reference numeral 501 denotes a color space conversion unit, which converts NTS-RGB pixel data to YC _r C _b pixel data in this embodiment. 502 is D
A CT processing unit, which outputs an output from the color space
DCT conversion processing is performed in units of 8 pixels. Reference numeral 503 denotes a quantization unit, and 8 × 8 = 64 Ds from the DCT transform unit 502
The CT coefficient data is quantized with coefficients having different weights to perform data reduction. 504 is a Huffman encoding unit;
The data quantized by the quantization unit 503 is represented by one D
The DC part is divided into a C part and 63 AC parts, and the DC part obtains a difference from the previously processed DC part of 8 × 8 pixels and performs Huffman coding. The AC section sorts the data into zigzag scans,
Two-dimensional Huffman coding is performed on the run length of “0” and the data that has appeared.

The operation of the data compression apparatus having the above configuration will be described with reference to FIGS. In the following description, it is assumed that the color data one pixel before stored in the latch 402 of FIG. 4 is x [i-1], and the color data stored in the latch 403 of FIG. 4 is PRE. Further, the pixel data output from the pixel data rearranging unit 401 in FIG.
[I]. x [i], x [i-1], and PRE are N
It is data in TSC-RGB format, and is 1 × 3 array structure data having R, G, and B data.

FIGS. 6 to 10 are flowcharts showing the processing procedure by the data compression device. First, step S6
At 0, the initialization of the latch 402 and the latch 403 is performed. That is, initial values are set to x [i-1] and PRE. Specifically, it is assumed that a normal image has a lot of black on a white background, and the initial value is x [i−1] = ｛255, 255, 25
5} (white), PRE = {0, 0, 0} (black). Next, in step S61, a unit of 8 × 8 pixels (hereinafter, referred to as a block) in the entire image is cut out, and the pixel data is held. Then, in step S62, one block of data is input, an area determination process, which will be described in detail later, is performed, and “0” or “1” as a determination result is output. Next, in step S63, if the determination result is "1", the process proceeds to step S64, and lossless compression processing described later in detail is performed. If the determination result is "0", the process proceeds to step S65 to perform a known ADCT compression process. Thereafter, when each of the above-described compression processes is completed, the process proceeds to step S66, and the result is stored in the address of the memory 6 designated by the control unit 7. Then, in step S67, it is determined whether or not all the blocks have been processed. If any unprocessed blocks remain, the process returns to step S61, and the above process is repeated. Then, when all blocks have been completed, this processing is completed.

FIG. 7 is a flowchart showing a detailed processing procedure of the area determination processing (step S62) shown in FIG. First, in step S621, the counter num1 is reset to "0". Next, in step S622,
It is checked whether or not there are four or more different colors in a line (eight pixels) by the processing shown in FIG. 8 described later, and the result is set or reset in a flag. In step S623, the result is determined based on the flag. If the flag is set, the process proceeds to step S624, and the counter num1 is incremented.

On the other hand, if the flag has not been set, the process directly proceeds to step S625, and all 8 in one block are processed.
It is determined whether the line has ended. Here, if NO, the process returns to step S622 and the above process is repeated. However, if YES, the process exits the loop, advances the process to step S626, and as a result of ending eight lines, determines whether the counter num1 is equal to or larger than the threshold value 4. As a result, in the case of YES, the process proceeds to step S627,
"1" is output, and if NO, step S628
And outputs "0".

FIGS. 8 and 9 are flowcharts showing the detailed processing procedure of the processing (step S622) shown in FIG. First, in step S62201, a buffer buf for storing color data for the first pixel in one line (eight pixels)
Substitute [0]. In step S62202, the counter num2 is initialized to "1". The reason why "1" is used instead of "0" is that one or more colors always exist in the line. Next, in step S62203, the remaining pixels (seven pixels) of one line are sequentially assigned to the color data buffer color. Then, step S622
In step 04, the counter i is reset, and
In step 2205, the flag is reset. In step S62206, color is compared with the color data of buf (i).
The process proceeds to step 7, and a flag is set because the color data already exists in the buf group. If different, the process directly proceeds to step S62208, and the counter i
Is incremented.

Next, in step S62209, i becomes n
It is determined whether it is less than um2, that is, whether or not all comparisons with the color data assigned to the buf group have been completed. If the determination is YES, the process returns to step S62206. If the determination is NO, the process exits the loop and proceeds to step S62210. In this step S622010, the flag flag is set.
Is "0" (that is, the same data does not exist in the buf group), and if it exists (NO), the
The process proceeds to 62213. However, if it does not exist (YES), the process proceeds to step S62211,
Substitute color for buf [num2]. In a succeeding step S62212, the counter num2 is incremented. Then, in step S62213, num
It is determined whether 2 is greater than or equal to a threshold. Here, in the case of YES, there is no need to process any more pixels, so the processing exits from the loop and proceeds to step S62216, where “1” is set.
Is output. On the other hand, if NO, step S6221
The process proceeds to 4 to determine whether all eight pixels in one line have been completed. As a result, if the result is NO, the process returns to the above-described step S62203. If the result is YES, the process exits the loop and proceeds to the step S62215, and outputs “0” assuming that there is no different pixel of the line or more than the threshold value I do.

FIG. 10 is a flowchart showing a detailed processing procedure of the reversible compression processing (step S64) shown in FIG. First, in step S901, the counter i is initialized, and in step S902, the pixels in the block are rearranged and sequentially processed. For this sorting method, for example, the order shown in FIG. 11 can be considered. As a result of the simulation, the rearrangement method (c) shown in FIG. 11 has the best compression ratio. In (a) and (b) above,
Non-adjacent pixels are processed for every eight pixels, so that the number of color data change points is increased and the compression efficiency is reduced.
Further, it is more efficient to match the line of the area determination unit with the processing line of the lossless compression unit. This is because, for example, FIG.
In the case of processing a block having such a pixel value as described above, if the area determination unit processes the eight lines shown in FIG. 13A, the area determination result is “0”. Therefore, if the reordering method shown in FIG. 11B is adopted in the reversible compression unit, the color data changes every time, so that the processed data increases.

However, if processing is performed in the sequence shown in FIG. 11A, only eight color data change points are required, and a high compression ratio can be obtained. Therefore, when (b) shown in FIG. 13 is adopted by the area determination unit, it is preferable that (b) or (c) shown in FIG. 11 is adopted by the lossless compression unit. Similarly,
When (a) shown in FIG. 13 is adopted by the area determination unit, it is preferable that (a) shown in FIG. 11 is adopted by the lossless compression unit. In the embodiment, since the lossless compression unit adopts (c) shown in FIG. 11 which has the best compression efficiency, the region determination unit uses FIG.
(B) shown in FIG.

Next, in step S903, x [i] and x [i-1] are compared, and if they are equal, the process proceeds to step S904 to output "0". However, if they are different, the process proceeds to step S905, outputs "1", and in step S906, x [i] is compared with PRE. Here, if x [i] and PRE are equal, the process proceeds to step S907 to output “0”. However, if different, the process proceeds to step S908, and outputs “1”. Then, in step S909, x [i] is output because x [i] is different from x [i-1] and PRE. next,
In step S910, x [i-1] is substituted for PRE. In step S911, the counter i is incremented, and in the next step S912, it is determined whether all pixels have been completed. When the processing is completed for all the pixels, the process proceeds to step S913, where x [0] (that is, x [i-1] of the pixel to be processed first in the next block) is used for processing the next block. [64] is assigned.

According to the embodiment described above, holding means for holding input pixel data from an original image in units of m × n in the horizontal and vertical directions, and inputting the data of the m × n pixels, Area determining means for outputting a control code based on a parameter obtained from the number of colors of a certain color; at least two different compressing means; and selecting means for selecting one from different compressing means according to the control code from the area determining means. , Appropriate compression can be performed on each area. In particular, lossless compression processing of computer-generated images containing high-frequency components, and irreversible compression processing of low-frequency images, such as gradation and natural images, whose compression efficiency does not increase and deterioration due to data changes is not noticeable, are performed. In addition, it is possible to perform compression having both image quality and compression efficiency while maintaining the outline and unity of color of an image created by a computer, and having good compression efficiency of a gradation image.

In the area determination method, the number of different colors in m × n pixels is counted, and the number is determined based on a threshold value, thereby separating a high-frequency part created by a computer and a low-frequency part such as a gradation / natural image. can do. Further, a means for counting the number of different pixels in one column in m × n pixels as a region determining means, and the result is k (k <k <
m), a means for counting how many rows exist, and a signal 1 when the result is greater than l (l <n) and a signal 2 when less than l, reduces the hardware scale of the area determination circuit. Areas can be separated with the same precision as the method of counting all colors in m × n while keeping it to a minimum. The compression efficiency can be improved by aligning the direction of the column provided at that time with the direction of rearranging the pixel data of the lossless compression method from parallel to serial. Further, by performing zigzag rearrangement, the number of color change points can be reduced, and the compression efficiency is improved.

<Modification> In the embodiment, one block is divided into eight lines and the number of pixels is determined by the area determination unit 3. However, the number of different color data in 64 pixels is counted, and the threshold value is set. It may be determined and determined. In that case, a determination result that does not depend on the direction of the gradation can be obtained.

Although the reversible compression section 4 is composed of two latches as shown in FIG. 4, it may be composed of two or more latches as shown in FIG. In that case, the signal corresponding to the signal 4b shown in FIG. 4 is 2 bits. However,
Since the output of the color data 13c (24 bits) shown in FIG. 14 is reduced, the final compression ratio is improved. Further, in the reversible compression section 4, the color data is output as it is with 24-bit RGB data, but the palette data may be separately provided, and the appearance data may be sequentially palletized or addressed. For example, if the pallet address is 8 bits, 24 bits are required to represent the color data, but 8 bits are sufficient, so that the compression ratio is improved.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of one device. Needless to say, the present invention can be applied to a case where the present invention is achieved by supplying a program to a system or an apparatus.

[0026]

As described above, according to the present invention,
By selecting the desired compression method in accordance with the characteristics of each block to be compressed can be compressed in high image quality improves the compression efficiency of each block. In particular,
As an inverse compression method, blocks to be compressed are arranged in one dimension.
It is an algorithm that compresses each pixel and compresses it for each pixel.
Specifically, the pixel to be compressed is a pixel data that has been compressed earlier.
Presumes compression based on whether data matches
Then, each block can be efficiently compressed. Furthermore,
According to claim 2 of the present invention, a lossless compression method and an irreversible compression method
In the selection control of the lossless compression method, each block is
Or one-dimensional array in the vertical direction and then compressing
And the selection can be performed with high accuracy.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of a data compression device according to an embodiment.

FIG. 2 is a block diagram showing a detailed configuration of an area determining unit 3 shown in FIG.

FIG. 3 is a block diagram showing a detailed configuration of a data processing circuit 202 shown in FIG.

FIG. 4 is a block diagram showing a detailed configuration of a lossless compression unit 4 shown in FIG.

FIG. 5 is a block diagram showing a detailed configuration of an ADCT processing unit 5 shown in FIG.

FIG. 6 is a flowchart showing an overall operation in the embodiment.

FIG. 7 is a flowchart illustrating a detailed processing procedure of the area determination processing illustrated in FIG. 6;

FIG. 8 is a flowchart illustrating a detailed processing procedure of the check processing illustrated in FIG. 7;

FIG. 9 is a flowchart illustrating a detailed processing procedure of the check processing illustrated in FIG. 7;

FIG. 10 is a flowchart showing a detailed processing procedure of the lossless compression processing shown in FIG. 6;

FIG. 11 is a diagram illustrating a reversible compression unit rearrangement method in the embodiment.

FIG. 12 is a diagram showing pixel data in an example.

FIG. 13 is a diagram illustrating a method of dividing an area determination unit into columns according to the embodiment.

FIG. 14 is a block diagram illustrating a configuration of a lossless compression unit according to a modification.

Claims (7)

(57) [Claims] An image data compression apparatus for compressing image data for each block of a plurality of pixel data , wherein the plurality of pixel data in the block is arranged one-dimensionally.
The raw data sequence is sequentially set as pixel data to be compressed, and
Compare pixel data with previously compressed pixel data
Reversible compression means for performing reversible compression based on the result obtained, the reversible compression means using a different algorithm
Lossy compression method for irreversibly compressing image data block by block
A row and an image in which a plurality of pixel data in the block is one-dimensionally arranged.
For the raw data string, a different image
Detects the number of raw data
Compressing each block based on the detection result
Means or means for compressing the irreversible compression means.
And a control means for controlling the image data compression. 2. The method according to claim 1, wherein the lossless compression means compresses the pixel data at the time of compression.
The direction of the one-dimensional array of data rows and the control means.
The image data compression device according to claim 1, wherein the direction of the one-dimensional array of the pixel data strings is the same direction . 3. The image processing apparatus according to claim 2, wherein the lossless compression unit includes a compression target pixel.
Whether the data matches the pixel data that was just compressed
The image data compression apparatus according to claim 1, wherein the image data is compressed based on the result . Wherein said lossless compression means, whether or not the compressed pixel data in the case does not match the pixel data compressed immediately before, are consistent with the pixel data compressed the compressed pixel data of two pixels before The image data compression apparatus according to claim 3, wherein the image data is compressed based on the result. Wherein said lossless compression means, wherein when the compressed pixel data does not match the pixel data compressed before immediately before and 2 pixels, to output the compressed pixel data as part of the compressed data 5. The method according to claim 4, wherein
3. The image data compression device according to claim 1. 6. The compression method according to claim 1, wherein the irreversible compression means performs compression by sequentially applying DCT transform, quantization, and Huffman coding to the pixel data group to be compressed. Image data compression device. 7. An image data compression method for compressing image data for each block of a plurality of pixel data, wherein a pixel data sequence in which a plurality of pixel data in the block is arranged one-dimensionally is sequentially set as pixel data to be compressed. A lossless compression step of performing reversible compression based on a result of comparing the pixel data to be compressed with pixel data that has been compressed earlier, and irreversible compression of the image data for each block using an algorithm different from the lossless compression step Irreversible compression step, and for a pixel data string in which a plurality of pixel data in the block are arranged one-dimensionally, the number of different pixel data present for each predetermined number of pixel data is detected, and the detection result corresponding to each block is detected. A control step of controlling each block to be compressed in one of the lossless compression step or the irreversible compression step. Image data compression method which is characterized by having.