JPH05292328A - Still picture coder - Google Patents

Abstract

PURPOSE:To eliminate the need for a large capacity field memory matching the input picture data at decoding by using a marker code and an information addition bit and decoding only optional coded data. CONSTITUTION:An optional marker code and an information addition bit describing its number are inserted in coded data for each code data relating to a picture border desired to be divided in input picture data from a still picture input means 1 at coding are inserted into coded data by a marker code insertion detection means 6 and an information addition bit insert detection means 8. A central processing unit 7 retrieves coded data desired to be decoded based on the information inputted from the means 6, 8 at decoding and only the coded data block is decoded. The means 8 may be a hardware or a software.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a still picture coding apparatus using a marker code indicating a delimiter of coded data in still picture coded data.

[0002]

2. Description of the Related Art In recent years, in addition to the conventional still image encoding system for monochrome binary images, research and development of still image encoding systems dealing with black and white halftones and color images have been actively conducted and standardization has been advanced. There is. Joint Photographic, which is an international standard of still image coding method for natural images
Express Group (hereinafter referred to as JPEG standard)
In order to efficiently compress and decompress enormous amounts of data in black and white halftone and color still images, the coding algorithm that uses a combination of discrete cosine coding, quantization and Huffman coding is used to specify the coding algorithm. It is carried out.

Hereinafter, description will be made using the still image coding method specified by the JPEG standard. However, if the coding method uses code data indicating a division of coded data equivalent to the marker code described below, it is possible. In particular, it is not limited to the JPEG standard.

In the JPEG standard, a data format and a marker code are specified in order to solve the problem of compatibility of encoded image data.

The data format defines the format of data when image data is encoded.

The marker code is a delimiter between the beginning and end of encoded data, the number / type of color elements used for color expression, or table information (quantization table It is code data used as a delimiter for adding decoding information such as a Huffman table) and is represented by 1 byte. There are two types of marker codes, one used alone and one having parameters. In the case of a marker code that requires a parameter, it is followed by 2 bytes of data indicating the length of the parameter and the required number of bytes of the parameter. In the JPEG standard, some marker codes that require this parameter can be used for their own purpose by applications not specifically shown in the application.

Next, a block diagram of a conventional still picture coding apparatus is shown in FIG. 1 is still image input means such as a scanner or camera for inputting still image data, 2 is a monitor for displaying still image decoded images, 3 is a field memory for storing decoded image data output to the monitor 2, 4 is the above A means 5 for compressing / decompressing the image data input from the still image input means 1 and a coded data memory 5 for storing the coded data compressed by the compression / decompression means 4. Reference numeral 6 inserts a marker code into the encoded data memory 5 at the time of encoding. Alternatively, it is means for detecting the marker code in the encoded data memory 5 at the time of decoding, and can be easily realized by either a software configuration or a hardware configuration. Reference numeral 7 denotes a central processing means, which inserts a marker code through the marker code insertion / detection means 6 as needed at the time of encoding, and inserts the marker code at the time of decoding.
The parameter change control of the compression / expansion means 4 is controlled according to various marcodes detected by the detection means 6.

FIG. 4 shows an embodiment in which a high-definition still picture is coded and decoded based on the block diagram of the conventional still picture coding apparatus as shown in FIG.

In FIG. 4, the still image input means 1 in FIG. 3 inputs 1280 dots in the horizontal direction and 800 dots in the vertical direction, and each dot inputs 8-bit (256 gradations) black and white high definition still image data. It is assumed that the resolution is 640 dots in the horizontal direction and 400 dots in the vertical direction on the monitor. The input image data has a capacity of 1.024 (megabytes).

Since the encoding is performed based on the JPEG standard, the input image data is 8 dots in the horizontal direction ×
It is assumed that the block data is encoded in block units of 8 dots in the vertical direction, and the block data can be input in block units from the still image input unit 1.

FIG. 4 will be described below with reference to FIG. 4a shows the input image data structure, FIG. 4b shows the structure of the coded data stored in the coded data memory 5 of FIG. 3, and FIG. 4c shows the structure necessary for outputting the decoded image of the high-definition still image to the monitor. 4 shows the capacity of the field memory 3 of FIG.

As described above, in the JPEG standard, the input image data is two-dimensionally encoded with a block of 8 dots in the horizontal direction × 8 dots in the vertical direction as a minimum unit, and therefore, in FIG. The input image data is 1600 to 16000.
It is divided into 0 blocks.

At the time of encoding, the input image data shown in FIG. 4a is encoded in block units by the compression / expansion means 4 of FIG. Then, as shown in the coded data structure of FIG.
The continuous encoded data is written in the encoded data memory 5 of FIG. In FIG. 4b, SOI (start of image) and EOI (end of image) indicate marker codes added to the beginning and end of the image data defined by the JPEG standard.

At the time of decoding, the encoded data of blocks [1] to [16000] similarly sandwiched between the marker code SOI and EOI shown in FIG. 4b is processed by the compression / expansion means 4 of FIG. It is continuously decoded and written in the field memory 3. The required field memory capacity at this time is shown in FIG. 4c. As shown in the figure, the field memory 3 needs to have a capacity equal to or larger than the input image data. That is, as described above, since the input image data has a capacity of 1.024 megabytes, the field memory requires a capacity of 1.024 megabytes or more.

Here, the coded data based on the JPEG standard is variable-length coded data, which is 8 dots ×
The amount of encoded data in block units of 8 dots varies depending on the image pattern, so the amount of data in each block is not constant. Therefore, it is practically difficult to search the encoded data to find the encoded data of an arbitrary block. Therefore, in the conventional encoded data structure, it is necessary to continuously decode the encoded data sandwiched between the marker code SOI and the EOI at the time of decoding.

The monitor resolution is 640 dots × 4.
Since it is 00 dots, the image data on the field memory is cut out by 640 dots × 400 dots and scrolled to display a still image in a wide range on the monitor.

In the above-mentioned conventional example, a black-and-white halftone still image is taken up for explanation, but the same can be said for a color still image, so it is not shown here.

In FIG. 4, marker codes other than the SOI and EOI defined by the JPEG standard may be described as needed, but they are omitted here.

[0019]

As described above, since the encoded data cannot be partially decoded by the conventional configuration, the encoded data of one screen of the input image data is continuously decoded. There was a need. For this reason, when the capacity of the field memory mounted on the still image encoding device is small, only a small range of image data can be input. Further, in order to input and display a high-definition still image over a wide range, a large-capacity field memory proportional to the input image data is required, which is an expensive device.

[0020]

According to the present invention, in order to solve the above problems, any one of the coded data corresponding to an image boundary to be divided in input still image data is defined by the JPEG standard or the like. The problem can be solved by adding an information addition bit indicating the number of the marker code to one marker code, and providing a means capable of separating / detecting any encoded data sandwiched by the marker code.

[0021]

With the above configuration, only the encoded data delimited by the marker code can be decoded. Therefore, even when a wide range of high-definition still image data is input and encoded,
Only the coded data blocks required according to the resolution of the monitor can be decoded. Therefore, it is not necessary to store the decoded image data irrelevant to the display in the field memory, and the field memory can be suppressed to the minimum capacity.

[0022]

1 is a block diagram of a still picture coding apparatus according to an embodiment of the present invention. In the figure, blocks having the same numbers as those in FIG. 3 have the same functions, and a description thereof will be omitted.

Reference numeral 8 denotes an information-added bit inserting / detecting means of the present invention, which is used in combination with the marker code inserting / detecting means 6.

That is, at the time of encoding, the still image input means 1
From the input image data sent from the device, for each coded data corresponding to an image boundary to be divided, an information addition bit indicating any one marker code and its number is inserted into the marker code insertion / detection means 5 and the information addition bit. -It is inserted into the encoded data by the detecting means 8.

At the time of decoding, the central processing unit 7 searches for encoded data to be decoded, based on the information input from the marker code inserting / detecting means 6 and the information-added bit inserting / detecting means 8. Then, only the encoded data block is decoded. The information-added bit inserting / detecting means 8 shown here can be easily realized by a software configuration or a hardware configuration.

Next, with reference to FIG. 1, one embodiment of the present invention will be specifically described with reference to FIG.

In FIG. 2, the input still image data and the resolution of the display monitor are all the same as in FIG. Therefore, the input image data is black and white halftone and has a capacity of 1.024 megabytes.

FIG. 2a shows the input image data structure, FIG. 2b shows the structure of the coded data stored in the coded data memory 5 of FIG. 1, and FIG. 2c outputs the decoded image of the high definition still image to the monitor. Indicates the required field memory capacity.

At the time of encoding, as shown in FIG. 2a, encoding is performed in order from the block [1] by the compression / decompression means 4 of FIG. 1. Here, the image boundary is set in units of 80 blocks. Then, the encoding is performed. That is, block [80]
, And block [81], between block [160] and block [161], ..., Between block [15920] and block [15921] are set as image boundaries and encoding is performed. As shown in FIG. 2b, the coded data memory 5 has a J memory for separating the coded data for each image boundary.
An arbitrary one marker code defined by the PEG standard and an information addition bit indicating the number of the marker code are inserted between coded data corresponding to an image boundary.

At this time, the marker code inserting / detecting means 6 and the information-added bit inserting / detecting means 8 shown in FIG.
use.

The marker code used here is not particularly limited. As described above, some marker codes defined by the JPEG standard have parameters and parameters do not have parameters. However, it is configured by combining an arbitrary one marker code and an information addition bit indicating the number of the marker code. If so, even if a parameter is inserted between the marker code and the information addition bit, which is not shown in FIG. 2b,
You can use several types of marker codes.

Also, as shown in FIG. 2b, since 16,000 block data are separated into 80 blocks,
The number of the marker code to write the information-added bit must be 0 to 199. This is 8 bits (1 byte)
Can be expressed as

Next, at the time of decoding, the resolution of the monitor 64
In order to decode in accordance with 0 dots × 400 dots, for example, in FIG. 2b, the marker code SOI and the marker are used based on the information detected by the marker code insertion / detection means 6 and the information addition bit insertion / detection means 8 in FIG. Coded data of block [1] to block [80] sandwiched between information addition bits 0 detected by the code, and block [161] to block sandwiched between information addition bit 1 and information addition bit 2 The encoded data of [240] is decoded. Similarly, blocks [321] to [4]
00], ..., And by decoding the encoded data from block [7841] to block [7920] in which the number of dots in the vertical direction of the decoded image data is 400 dots,
The decoded image data required for full-screen display on the monitor can be obtained.

Similarly, while detecting the marker code and the information addition bit of the encoded data, the block [808
1] to block [8160] coded data, block [8241] to block [8320] coded data, ..., Block [15921] to block [16]
Decoding image data necessary for full-screen display on the monitor can be obtained by performing decoding according to the coded data of [000].

In other words, the still image input image data in a wide range is provided with the image boundary as described above, and encoding and decoding are performed using the means of the present invention, so that the entire surface of the monitor is displayed in the horizontal direction by 2. Decoded image data can be created by dividing it into 50 stages in the vertical direction.

Further, it is possible to make the decoding range of the decoded data more finely by providing the image boundaries finer than the above-mentioned image boundaries and increasing the number of marker codes and information addition bits. Therefore, the image boundary needs to be optimally set according to the specifications of the still image encoding device.

In the above-described embodiment of the present invention, the black-and-white halftone still image is taken up for explanation, but it is obvious that the same can be said for the color still image, and it is not particularly shown here.

[0038]

As described above, according to the present invention, it is possible to decode only arbitrary coded data by using the marker code and the information addition bit as described above. Therefore, a wide range of high-definition still image data can be input. Even in the case of encoding, only the image data necessary for monitor display can be decoded, and therefore a large-capacity field memory matching the input image data is not required at the time of decoding. Therefore, the capacity of the field memory can be configured with the minimum capacity required for monitor display.

Since the image data range to be written in the field memory at the time of decoding can be changed according to the image boundary position provided at the time of encoding, scroll display can be performed by changing the sequentially decoded image data area. Even input image data that exceeds the field memory capacity can be reproduced with high precision over the entire range.

[Brief description of drawings]

FIG. 1 is a block diagram of a still image encoding device according to an embodiment of the present invention.

FIG. 2 is a data configuration diagram in the still image encoding device of the present invention (a) input image data configuration diagram (b) encoded data configuration diagram (c) diagram showing required field memory capacity

FIG. 3 is a block diagram of a conventional still image encoding device.

FIG. 4 is a data configuration diagram of a conventional still image encoding device. (A) Input image data configuration diagram (b) Encoded data configuration diagram (c) Diagram showing required file memory capacity

[Explanation of symbols]

 8 Information addition bit insertion / detection means

Claims (1)

[Claims] 1. In still picture coding, an arbitrary one marker code indicating a delimiter of coded data and an information-added bit indicating the order of the marker code are inserted into the coded data, and a still picture is inserted. An information addition bit insertion / detection means for detecting the marker code and the information addition bit inserted in the encoded data at the time of image decoding is provided, and the marker code is referred to with reference to the information addition bit. A still image encoding device capable of arbitrarily selecting and decoding one of a plurality of encoded data blocks separated by.