JP2007201537A - Image compression method, and information processing apparatus provided with image compression means - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve problems of adopting the "Markov coding method" such as the JBIG in the case of applying compression processing to attribute data in 2 to 4 bits, an image with a limited gradation, and a dither image for expressing halftone by using a plurality of pixels that a proper reference range cannot be selected as bits of data are greater and capacity of a line buffer is increased. <P>SOLUTION: In an image compression method disclosed herein wherein image data of a coding object coded by referencing image data around the coding object image data and using the correlation between the referenced image data and the object image data, control for changing referred pixels depending on the arrangement of the image data is performed. For example, when the image data are arranged in a lateral direction, a region in a laterally long direction is adopted for a region to be referenced for the compression. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a compression processing method that refers to a finite number of image data processed before a certain pixel when predicting a certain pixel, and in the case of multi-bit data, refers to the referred pixel based on the arrangement of the image data. The present invention relates to a function for determining a pixel.

  In recent years, functions of PCs (personal computers) and many other information processing apparatuses have been enhanced. At the same time, multimediaization has progressed, and large data such as image data, audio data, and moving image data are handled. As a result, the data handled by PCs has grown. In this situation, a technique for compressing data for data transmission and storage is required, and various compression techniques have been advanced.

  As compression methods, there are lossless encoding (information storage type encoding) and lossy encoding (information nonconservation type encoding). In general, lossless encoding is mainly used when compressing a binary image, and irreversible encoding is used for a natural image having a large number of gradations.

  As a prominent method of the lossless compression method, there is a Markov model encoding method. This principle is as follows. In image data, the correlation between adjacent pixel data is strong. That is, it can be considered that the probability distribution of what the image data of a certain pixel (pixel of interest) is depends on a finite number of pixel data processed before that. Based on this premise, when a certain pixel is predicted, the prediction is performed using pixel data processed before that. By transmitting and storing this information, the entropy of the image data can be reduced.

  Figure 1 shows the reference pixel region of JBIG (Joint Bi-level Image coding experts Group) that performs compression using the principle of Markov model coding. JBIG is a typical example of lossless encoding. Since JBIG assumes binary data, the pixel shown in FIG. 1 is used as a reference range. In this Markov model encoding method, the compression performance varies greatly depending on the reference pixel used. Accordingly, a pixel arranged near the pixel of interest has a stronger correlation with the pixel of interest, so a neighboring pixel is selected.

  There is JPEG as a lossy compression method. The JPEG compression method is as follows. The input image is made into an 8 × 8 pixel block for each color component such as RGB. This block is subjected to DCT transform (discrete cosine transform). The converted value is divided by an 8 × 8 quantization table into an integer value. A high-frequency AC component in a normal image is hardly recognized by human eyes. Therefore, this quantization table decreases the coefficient corresponding to the low frequency component and increases the coefficient corresponding to the high frequency component. Then, the low frequency component can be stored, and the high frequency component can be almost zero. By increasing the value of 0, the compression efficiency becomes higher when Huffman coding is performed thereafter.

In the conventional technology, JBIG is suitable for binary images and JPEG is suitable for natural images as a compression method. When transferring images over multiple pages, binary images are compressed with JBIG. Some multi-valued image pages have been proposed to have a control means for compressing with JPEG (see Patent Document 1).
JP-A-6-54203

  As described above, JBIG is often used for binary images, and JPEG is often used for natural images. However, there is no appropriate method in the case of attribute data such as 2 to 4 bits, a limited gradation image, and a tether image that represents a halftone using a plurality of pixels.

  Also, when using the “Markov model encoding method” such as JBIG, the more the data becomes multi-bit, the more the reference range cannot be selected, and the line buffer becomes larger.

  An object of the present invention is to provide an effective method for the compression of image data as described above.

The invention according to claim 1 is an image compression method for encoding image data to be encoded with reference to surrounding image data, and encoding them according to their correlation.
An image compression method is characterized in that the reference area is changed in accordance with the arrangement of the image data.

The invention according to claim 2 is the image compression method,
2. The image compression method according to claim 1, wherein a change is made so that the reference area is wide in the direction parallel to the bit arrangement of the image data.

The invention according to claim 3 is an image compression method using JBIG.
3. The image compression method according to claim 2, wherein a 3-line template is used when compressing 1-bit data, and a 2-line template is used when compressing multi-bit data.

The invention according to claim 4 is an image compression method using JBIG, in multi-bit processing,
4. The image compression method according to claim 1, wherein the AT position is arranged in a peer bit of the reference pixel.

According to a fifth aspect of the present invention, in an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to a correlation therebetween.
An information processing apparatus comprising image compression means for realizing the image compression method according to claim 1.

The invention according to claim 6 is the image compression method,
2. The image compression method according to claim 1, wherein when a certain bit in the pixel of interest is compressed, a peer bit in the reference pixel is used as reference image data.

The invention according to claim 7 is an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to a correlation therebetween.
An information processing apparatus comprising image compression means for realizing the image compression method according to claim 6.

The invention according to claim 8 is an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to their correlation.
8. The information processing apparatus according to claim 7, further comprising bit division means for dividing the image data into bits for compression bit by bit.

The invention according to claim 9 is an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to a correlation therebetween.
8. The information processing apparatus according to claim 7, further comprising: a bit combining unit that combines the expanded data expanded for each bit to restore the image data.

According to a tenth aspect of the present invention, there is provided an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to a correlation therebetween.
A bit dividing means for dividing the image data into bits for compression bit by bit;
Bit combining means for combining the expanded data expanded for each bit to restore the image data;
The information processing apparatus according to claim 7, further comprising:

  According to the present invention, when processing image data to be encoded, reference is made to reference pixels according to the arrangement of the image data in an image compression method that refers to the surrounding image data and encodes them according to their correlation. Control to change.

  With this control, the compression rate can be increased, and a line buffer that normally requires a larger capacity for multi-bit data can be reduced.

  The best mode for carrying out the invention is the following embodiment. In this specification, JBIG is taken as an example, but even if the compression means is different, if the reference area is determined by the arrangement of the image data, it is included in the scope of the present invention.

<Method using standard JBIG>
The present invention is an effective method when image data to be compressed is 2 to 4 bits and standard JBIG is used. The present invention can be implemented without changing the standard JBIG circuit.

  FIG. 2 shows an arrangement of 2-bit image data in the embodiment of the present invention.

  FIG. 3 shows an arrangement of 3-bit image data in the embodiment of the present invention.

  FIG. 4 shows an arrangement of 4-bit image data in the embodiment of the present invention.

  This 2-4 bit image data is expressed as one pixel at the time of output.

  In this specification, several bits of data expressed as one pixel at the time of output are called pixel data. The data of each bit constituting the pixel data is called image data.

  The direction of processing is as follows.

  The horizontal direction is processed from left to right as indicated by arrows. This direction is called the j direction as shown in the figure. Or it is called the bit string direction.

  The vertical direction is processed from top to bottom as indicated by the arrows. This direction is called the i direction as shown in the figure. Or called the line direction.

  For the processing direction of all the images, first, the column that is the starting point of the i column is selected and the processing is performed in the j direction. When this column is completed, the column next to the i column is selected, and the process is performed again in the j direction. In this way, processing is performed from the upper left pixel to the lower right pixel.

  In this specification, the bit arrangement is the arrangement shown in FIGS. 2 to 4 and performed in the processing order as described above. However, even when the bit arrangement is different and the bit arrangement is performed according to the arrangement, the image data If the reference region is determined by the sequence, it is included in the present invention.

  Next, reference pixels of JBIG at the time of 1 to 4 bits in this embodiment are shown.

(1) For 1-bit data The reference area of the 3-line template and 2-line template in the standard JBIG is as shown in FIG. The 3-line template refers to a pixel close to the target pixel at the time of 1-bit data. The 2-line template refers to a pixel close to the target pixel while reducing the line buffer.

(2) At the time of 2-bit data FIG. 5 shows reference image data at the time of the standard JBIG 3-line template and 2-line template when the first bit is the image data of interest in the 2-bit data.

  In this case, in the line direction, the 3-line template refers to the previous line and the previous two-line pixel, but the 2-line template refers to the immediately previous line, but the previous 2-line pixel I can't refer to it.

  On the other hand, in the bit string direction, in the 3-line template, only the peer bit of the previous pixel data (in this case, the first bit) is referred to, and in the 2-line template, simultaneously with referring to the one pixel, It refers to even the same bit of data.

  FIG. 6 shows the reference image data in the standard JBIG 3-line template and 2-line template when the second bit is the target image data.

  In this case as well, in the line direction, the 3-line template refers to the previous line and the pixel 2 lines before, but the 2-line template refers to the immediately previous line, but 2 lines. The previous pixel cannot be referenced.

  On the other hand, in the bit string direction, in the 3-line template, only the data of the previous pixel is referred to, and in the 2-line template, the previous pixel is referred to at the same time as the previous pixel.

  In this 2-bit data, whether the compression rate is higher when the line direction is referred to more or when the bit string direction is referenced is determined by the characteristics of the image. I cannot judge which is more advantageous. However, in the 2-line template, the line buffer for one line can be reduced as compared with the 3-line template. If the device to be mounted is a device that does not process 1-bit data, the 2-line template can be mounted with fewer line buffers.

(3) At the time of 3-bit data FIG. 7 shows the reference image data at the time of the standard JBIG 3-line template and 2-line template when the first bit is image data of interest in the 3-bit data.

  In this case, in the line direction, the three-line template refers to the immediately preceding line and the pixel two lines before, but the two-line template refers to only the immediately preceding line.

  On the other hand, in the bit string direction, the 3-line template does not refer to the data of the previous pixel, and the 2-line template refers to the first and second bits of the previous pixel.

  FIG. 8 shows the reference image data in the standard JBIG 3-line template and 2-line template when the second bit is the target image data in the 3-bit data.

  In this case, in the line direction, the 3-line template refers to the immediately preceding line and the pixel 2 lines before, but the 2-line template refers to only the immediately preceding line.

  On the other hand, in the bit string direction, only the first bit of the data one pixel before is referred to in the 3-line template, and all bits are referred to the previous pixel in the 2-line template.

  FIG. 9 shows the reference image data for the standard JBIG 3-line template and 2-line template when the third bit is the target image data.

  In this case, in the line direction, the 3-line template refers to the 2 or 3 bits of the previous line and the previous 2 pixels, but the 2-line template refers to only the previous line.

  On the other hand, in the bit string direction, in the 3-line template, only the 1st and 2nd bits of the data of the previous pixel are referred to. Refers to.

  In the case of the three-line template, although the pixel two lines before is referred to, when the first pixel is 1 or 3 bits, all the bits of the pixel cannot be referred to.

  On the other hand, in the case of the 2-line template, all the bits can be referred to the data one pixel before in the bit string direction.

  From the above points, when an arbitrary image is compressed, the 2-line template is more likely to have a better compression rate.

(4) At the time of 4-bit data FIG. 10 shows reference image data at the time of standard JBIG 3-line template and 2-line template when the first bit is image data of interest in 4-bit data.

  In this case, in the line direction, the 3-line template refers to the 1st to 3rd bits of the immediately preceding line and the 1st and 2nd bits of the pixel 2 lines before, but in the 2 line template, the immediately preceding line All bits are referred to for this pixel.

  On the other hand, in the bit string direction, the 3rd line template does not refer to the 4th bit of the pixel of interest, and the 2line template refers to the same bit before the 1st pixel (in this case, the 1st bit).

  FIG. 11 shows the reference image data in the standard JBIG 3-line template and 2-line template when the second bit is the target image data in the 4-bit data.

In this case, in the line direction, the 3-line template refers to all bits of the immediately preceding line and the 1st to 3rd bits of the pixel 2 lines before, but in the 2-line template, the pixels of the immediately preceding line are referred to. Only all bits are referenced.
On the other hand, in the bit string direction, the 3rd line template refers to the 3rd and 4th bits of the target pixel, and the 2 line template refers to the 1st and 2nd bits of the previous pixel.

  FIG. 12 shows the reference image data in the standard JBIG 3-line template and 2-line template when the third bit is the target image data in the 4-bit data.

  In this case, in the line direction, the 3-line template refers to all the bits of the immediately preceding line and the 2nd to 4th bits of the pixel 2 lines before, but in the 2-line template, the pixels of the immediately preceding line are referred to. Only all bits are referenced.

  On the other hand, in the bit string direction, the first bit of the previous pixel is referred to in the 3-line template, and the first and third bits of the previous pixel are referred to in the 2-line template.

  FIG. 13 shows standard JBIG 3-line template and 2-line template reference image data when the 4th bit is the target image data.

  In this case, in the line direction, the 3-line template refers to the 2nd to 4th bits of the immediately preceding line and the 3rd and 4th bits of the 2nd previous pixel. 2 to 4 bits are referenced.

  On the other hand, in the bit string direction, the 3rd line template refers to the 1st and 2nd bits of the previous pixel, and the 2nd line template refers to all the bits of the previous pixel.

  In the case of the 3-line template, the peer bit cannot often be referred to. On the other hand, in the case of the 2-line template, the peer bit of the target pixel can be referred to for the data one pixel before in the bit string direction.

  From the above points, when an arbitrary image is compressed, the 2-line template is more likely to have a better compression rate.

  When this method was used to compress 4-bit image data, it was improved from 5.47% with the 3-line template to 5.12% with the 2-line template.

  The following describes how much memory capacity can be reduced by reducing the line buffer for one line.

  The line buffer must hold image data for one line. Therefore, it is necessary to have as many bits as the width of the image to be processed.

  FIG. 14 shows the prescribed paper size. For example, when processing a 600ppi image with the A3 size, the line buffer size is 7,157 bits. This is the size of 1 bit per pixel. A line buffer of 14,173 bits for 2-bit data, 21,471 bits for 3-bit data, and 28,628 bits for 4-bit data is required.

  Of course, if the image data is large, a large line buffer is required. Therefore, the 2-line template does not need to have an extra line buffer, and is advantageous in terms of cost.

  In this way, by increasing the reference area in the direction parallel to the image data arrangement, the compression rate and the line buffer can be reduced, which is advantageous over normal compression.

  As a usage method, a 3-line template is used so that the compression rate is high when compressing 1-bit data, and the line buffer is small when compressing multi-bit data, and the compression rate is likely to be high. By using the two-line template, efficient compression can be performed.

  In the present embodiment, the AT is a standard position, but there is a method of improving the compression rate by arranging the AT in the same bit of adjacent pixel data.

<JBIG compression method for switching the reference area for each bit>
In this embodiment, 4-bit data is taken as an example.

  Similar to the first embodiment, the data array is considered to be the array shown in FIG.

  FIG. 15 shows reference image data for a 3-line template and 2-line template when the target image data is the first bit in the embodiment of the present invention.

  As shown in the figure, when the first bit of the target pixel is the target image data, the first bit of the peripheral pixels is used as the reference pixel.

FIG. 16 shows reference image data for a 3-line template and 2-line template when the target image data is the second bit in the embodiment of the present invention.
As shown in the figure, when the second bit of the target pixel is the target image data, the second bit of the peripheral pixels is used as a reference pixel.

FIG. 17 shows reference image data for a 3-line template and 2-line template when the target image data is the third bit in the embodiment of the present invention.
As shown in the figure, when the third bit of the target pixel is the target image data, the third bit of the peripheral pixel is used as a reference pixel.

  FIG. 18 shows reference image data for a 3-line template and 2-line template when the target image data is the 4th bit in the embodiment of the present invention.

  As shown in the figure, when the fourth bit of the target pixel is the target image data, the fourth bit of the peripheral pixel is used as a reference pixel.

  By dividing each bit in this way, it is possible to compress by referring to peer bits having a strong correlation.

  In general, pixel values often change slowly in an image, and in the case of a limited gradation image or a tethered image, the compression rate is particularly good with higher bits.

  FIG. 19 is a diagram illustrating an implementation method for compressing multi-bit image data according to the present embodiment.

  At the time of compression, bits are divided and each bit data is compressed.

  At the time of expansion, the respective bit data are expanded and the bits are combined.

  This operation realizes the present embodiment.

  FIG. 20 shows a configuration example for realizing the present embodiment when the compression process and the decompression process are performed by the same circuit. In JBIG, compression processing and decompression processing can be realized with the same circuit.

  Reference numeral 10 denotes an input control circuit that performs input control of image data. The input control circuit includes a bit division circuit 11. This bit division circuit functions during compression.

  Reference numeral 20 denotes an output control circuit that performs output control of image data. The output control circuit includes a bit combination circuit 21. This bit coupling circuit functions during expansion.

  Reference numerals 101 to 104 denote compression / decompression circuits. Each block performs compression / decompression processing for each bit.

  The operation in this configuration will be described.

  The following processing is performed during compression. Data is input to the input control circuit. At this time, the bit division circuit divides the input data into bits and transmits the data to each compression / decompression circuit. The data compressed by the compression / decompression circuit is passed to the output control unit. The output control unit holds each bit in the memory.

  When decompressing, the following processing is performed. Data compressed for each bit is input to the input control circuit and transmitted to each compression / decompression circuit. The data decompressed by the compression / decompression circuit is passed to the output control unit. In the output control unit, the bits are combined by the bit combination circuit to decode the data.

  FIG. 21 shows another implementation method of the present embodiment, in which one compression circuit recognizes the number of bits of data and the number of bits to be processed, and repeats compression processing for the number of bits. FIG.

  FIG. 21 shows an example of processing N-bit data.

  Processing during compression is performed as follows.

  In the first time, the first bit is selected, and the compression processing is performed while selecting the first bit by the selection circuit.

  In the second time, the second bit is selected, and the compression process is performed while the second bit is selected by the selection circuit.

  In the third time, the third bit is selected, and the compression processing is performed while the third bit is selected by the selection circuit.

  The above processing is repeated for each bit, and the Nth bit of the most significant bit is processed.

  In the Nth time, the Nth bit is selected, and compression processing is performed while the Nth bit is selected by the selection circuit.

  The decompression process is performed as follows.

  In the first time, compressed data of the first bit is selected and decompression processing is performed.

  The second time, the compressed data of the second bit is selected and decompression processing is performed.

  In the third time, compressed data of the third bit is selected and decompression processing is performed.

  The above processing is repeated for each bit, and the Nth bit of the most significant bit is processed.

  In the Nth time, compressed data of the Nth bit is selected and decompression processing is performed.

  When all bits of data are decoded, bit combination is performed to decode the data.

  In the above description, processing is performed from the lower bits, but the order is not limited.

  As described above, by setting the reference region in consideration of the bit arrangement and performing the compression process, an effect on the compression rate can be obtained.

<Markov model encoding method referring to 12-bit image data>
In this embodiment, an embodiment of a reference method in the case of referring to 21-bit image data by an encoding method using the Markov model encoding method will be described. In the present embodiment, the reference bits are 21 bits. However, any method that performs compression processing by determining a reference area based on the arrangement of image data regardless of the number of reference bits is included in the present invention.

  FIG. 22 shows a reference range when 1-bit data is compressed by an encoding method using the Markov model encoding method. In the image data, the closer to the target pixel, the stronger the correlation with the target pixel. Therefore, when compressing, the compression is performed with reference to a pixel close to the target pixel as shown in the figure.

  FIG. 23 shows a reference range when 2-bit data is compressed by the encoding method using the Markov model encoding method.

  FIG. 24 is a diagram showing the arrangement of reference ranges when the second bit of 2-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 25 is a diagram showing the arrangement of reference ranges when the first bit of 2-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 26 shows a reference range when 3-bit data is compressed by an encoding method using the Markov model encoding method.

  FIG. 27 is a diagram showing the arrangement of reference ranges when the third bit of 3-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 28 is a diagram showing the arrangement of reference ranges when the second bit of 3-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 29 is a diagram showing the arrangement of reference ranges when the first bit of 3-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 30 shows a reference range when 4-bit data is compressed by the encoding method using the Markov model encoding method.

  FIG. 31 is a diagram showing the arrangement of reference ranges when the fourth bit of 4-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 32 is a diagram showing the arrangement of reference ranges when the third bit of 4-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 33 is a diagram showing the arrangement of reference ranges when the second bit of 4-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  FIG. 34 is a diagram showing the arrangement of reference ranges when the first bit of 4-bit data is compressed. As shown in the figure, compression is performed with reference to pixel data close to the pixel including the target pixel data.

  As described above, by setting the reference region in consideration of the bit arrangement and performing the compression process, an effect on the compression rate can be obtained.

It is a figure which shows the reference pixel area | region of JBIG which performs compression using the principle of a Markov model encoding method. It is a figure which shows the arrangement | sequence of 2-bit image data in 1st Example. It is a figure which shows the arrangement | sequence of 3 bit image data in 1st Example. It is a figure which shows the arrangement | sequence of 4-bit image data in 1st Example. FIG. 6 is a diagram illustrating reference image data in a standard JBIG 3-line template and 2-line template when the first bit is image data of interest in 2-bit data in the first embodiment. FIG. 6 is a diagram showing reference image data in a standard JBIG 3-line template and 2-line template when the second bit is image data of interest in the 2-bit data in the first embodiment. FIG. 6 is a diagram showing reference image data when a standard JBIG 3-line template and 2-line template are used when the first bit is the target image data in the 3-bit data in the first embodiment. FIG. 5 is a diagram showing reference image data in the case of standard JBIG 3-line template and 2-line template when the second bit is the target image data in the 3-bit data in the first embodiment. V is a diagram showing reference image data in the case of the standard JBIG 3-line template and 2-line template when the third bit is the target image data in the 3-bit data in the first embodiment. FIG. 4 is a diagram showing reference image data in a standard JBIG 3-line template and 2-line template when the first bit is image data of interest in 4-bit data in the first embodiment. FIG. 6 is a diagram showing reference image data in the case of standard JBIG 3-line template and 2-line template when the second bit is image data of interest in the 4-bit data in the first embodiment. FIG. 6 is a diagram showing reference image data in a standard JBIG 3-line template and 2-line template when the third bit is image data of interest in the 4-bit data in the first embodiment. FIG. 4 is a diagram showing reference image data in the case of standard JBIG 3-line template and 2-line template when the fourth bit is image data of interest in the 4-bit data in the first embodiment. Indicates the prescribed paper size for estimating the capacity of the line buffer. It is a figure which shows the reference image data at the time of the 3 line template when the attention image data is the 1st bit, and the 2 line template in 2nd Example. It is a figure which shows the reference image data at the time of the 3 line template when the attention image data is the 2nd bit, and the 2 line template in 2nd Example. It is a figure which shows the reference image data at the time of the 3 line template when the image data of interest is the 3rd bit, and the 2 line template in 2nd Example. It is a figure which shows the reference image data at the time of the 3 line template when the image data of interest is the 4th bit, and the 2 line template in 2nd Example. It is a figure which shows the realization method for compressing multibit image data in 2nd Example. A configuration example in the case where the compression processing and the decompression processing are performed by the same circuit in the second embodiment will be described. FIG. 10 is a diagram illustrating a method of recognizing the number of bits of data and the number of bits to be processed by one compression circuit and repeating the compression process for the number of bits in the second embodiment. It is a figure which shows the reference image data at the time of 1 bit data in 3rd Example. It is a figure which shows the reference image data at the time of 2 bit data in 3rd Example. It is a figure which shows the reference image data in case the 2nd bit data of 2 bit data is a focused pixel in 3rd Example. It is a figure which shows reference image data when the data of the 1st bit of 2-bit data are a focused pixel in 3rd Example. It is a figure which shows the reference image data at the time of 3 bit data in 3rd Example. It is a figure which shows the reference image data when the 3rd bit data of 3 bit data is a focused pixel in 3rd Example. It is a figure which shows the reference image data when the data of the 2nd bit of 3 bits data is a focused pixel in 3rd Example. It is a figure which shows the reference image data when the data of the 1st bit of 3 bit data is a focused pixel in 3rd Example. It is a figure which shows the reference image data at the time of 4 bit data in 3rd Example. It is a figure which shows the reference image data in case the 4th bit data of 4 bit data are the attention pixel in 3rd Example. It is a figure which shows the reference image data when the 3rd bit data of 4 bit data are the attention pixel in 3rd Example. It is a figure which shows the reference image data in case the 2nd bit data of 4 bit data is a focused pixel in 3rd Example. It is a figure which shows reference image data when the data of the 1st bit of 4-bit data are the attention pixel in 3rd Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Input control part 11 Bit division | segmentation circuit 20 Output control part 21 Bit coupling circuit 101 Compression / decompression circuit 1
102 Compression / decompression circuit 2
103 Compression / decompression circuit 3
104 Compression / decompression circuit N

Claims (10)

In an image compression method for encoding image data to be encoded with reference to the surrounding image data and encoding them according to their correlation,
An image compression method characterized by changing a reference area according to an arrangement of image data. In the image compression method,
2. The image compression method according to claim 1, wherein a change is made so that the reference area is wide in a direction parallel to the bit arrangement of the image data. In the image compression method using JBIG,
3. The image compression method according to claim 2, wherein a 3-line template is used when compressing 1-bit data, and a 2-line template is used when compressing multi-bit data. In multi-bit processing with the image compression method using JBIG,
4. The image compression method according to claim 1, wherein the AT position is arranged in a peer bit of the reference pixel. In an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to their correlation,
An information processing apparatus comprising image compression means for realizing the image compression method according to claim 1. In the image compression method,
2. The image compression method according to claim 1, wherein when a certain bit in the pixel of interest is compressed, a peer bit in the reference pixel is used as reference image data. In an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to their correlation,
An information processing apparatus comprising image compression means for realizing the image compression method according to claim 6. In an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to their correlation,
8. The information processing apparatus according to claim 7, further comprising bit division means for dividing the image data into bits for compression bit by bit. In an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to their correlation,
8. The information processing apparatus according to claim 7, further comprising bit combining means for combining the expanded data expanded for each bit to restore the image data. In an information processing apparatus having an image compression function for encoding image data to be encoded with reference to surrounding image data and encoding them according to their correlation,
A bit dividing means for dividing the image data into bits for compression bit by bit;
Bit combining means for combining the expanded data expanded for each bit to restore the image data;
The information processing apparatus according to claim 7, further comprising:

Images