JP2002199226A - Image data compression device and image data compression method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image data compression device that can obtain a high compression rate even in two-dimensional processing for a pseudo gradation image while utilizing advantages of conventional technologies. SOLUTION: The image data compression device that compresses pseudo gradation image data consisting of pixels of two-dimensional arrangement is provided with a means that sets a two-dimensional block to detect a size of a pseudo gradation pattern from part or all of data of a compression object image, a means that divides the compression object image in the unit of the decided block and discriminates whether or not the data in the unit of the block have repetition, a means that codes the data in the unit of the block when the data in the unit of the block have no repetition, a means that counts number of the repetitions and codes the data when the data in the unit of the block have repetition, and a means that generates a compression code from the coded code.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an image data compression apparatus and an image data compression method for compressing pseudo gradation image data composed of two-dimensionally arranged pixels. In particular, it can be suitably used when image data is temporarily stored in a memory in a printer device.

[0002]

2. Description of the Related Art In recent years, with the increase in speed and reliability of laser recording type printers, a function of guaranteeing received data from data transfer to required ejection has been required. For this reason, a large-capacity memory is required in the printer device, but a data compression method with the highest possible speed and a high compression ratio is preferable in view of the cost of the device. Images composed of pixels in a two-dimensional array including data output by a printer are classified into binary images and multivalued images. A multi-valued image is usually used for a natural image, and its gradation level is often set to a power of 2 (16 gradations / 64 gradations / 128 gradations / 256 gradations). Binary images are further classified into character images and pseudo-gradation images. Photo data taken from a digital camera to a personal computer or data exchanged on a personal computer or by data communication is a multi-valued image, but when these data are transmitted / received by facsimile or output by a printer, they are converted into pseudo gradation data. There is a need.

[0003] The pseudo gradation image is a technique in which a grid, which cannot be distinguished by the naked eye, is painted in black and white in accordance with the gradation value so that the human eye has a gradation. The method and the error diffusion method are widely known.

[0004] As a compression method for a binary image, compression using simple run-length coding or MH (Modified Huffman) used for facsimile or the like is used.
n) A high compression rate can be obtained by the encoding method. However, it is difficult to obtain a high compression ratio by the simple run-length coding or the MH coding in the case of a pseudo gradation image having a short run-length length.

As a technique for improving the compression ratio of a pseudo gradation image, there is, for example, JP-A-9-65147. This technique focuses on the fact that, in the dither method, which is one of the pseudo gradation images, the same black and white pattern is repeated when the same gradation continues in the original image, and the repetition is encoded separately from the run length. By adopting the method, the compression ratio of the pseudo gradation image is greatly improved.

[0006]

However, the invention disclosed in Japanese Patent Application Laid-Open No. 9-65147 is effective when the block to be dithered is small (for example, 4 × 4: 16 gradations). As the size becomes larger, for example, as shown in FIG. 8, the repetition of the combination of black and white patterns is often lost depending on how dither is painted. The dither in FIG. 8 has 5 × 5 pixels. For example, in line A and line B, the black and white pattern is repeated in black 3 white 2 (5 pixel cycle), and the repetition can be detected by the method disclosed in Japanese Patent Application Laid-Open No. 9-65147. However, in line C, white 2 black 1
Since a pattern of white 1 black 1 (5 pixel cycle) is repeated, a method of detecting a combination of black and white patterns like this method cannot detect pattern repetition, and a high compression rate cannot be obtained. I will. The technique disclosed in Japanese Patent Application Laid-Open No. 9-65147 is a compression method in which image data is reduced in redundancy only in the main scanning direction, that is, one-dimensionally. If it is reduced, a further improvement in the compression ratio is expected.

[0007] As a conventional technique for solving this problem,
For example, there is Japanese Patent No. 2840420. This technology is
As shown in FIG. 9, for example, intermediate data composed of a pattern and a pattern run length is created by preprocessing for comparing data with preceding data in units of four lines, and is effective in the case of a binary image as shown in FIG. It is. However, with this technique, contrary to Japanese Patent Application Laid-Open No. 9-65147, an improvement in the compression ratio of the pseudo gradation image cannot be expected.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and it is possible to obtain a high compression ratio even in two-dimensional processing of a pseudo gradation image while taking advantage of the prior art. It is an object to provide an image data compression device and an image data compression method.

[0009]

In order to solve the above-mentioned problem, a first aspect of the present invention is an image data compression apparatus for compressing pseudo gradation image data composed of two-dimensionally arranged pixels. A means for detecting the size of the key pattern is provided.

According to a second aspect of the present invention, there is provided an image data compression apparatus for compressing pseudo-gradation image data composed of two-dimensionally arranged pixels, wherein a two-dimensional block is set and a part of the image to be compressed is set. Alternatively, a means for detecting the size of the pseudo gradation pattern from all data, a block unit for performing encoding is determined from the detected size of the pseudo gradation pattern, and the compression target image is divided into the determined block units. A means for determining whether or not the data in the block has repetition, a means for encoding the data in the block when the data in the block does not have a repetition, and a means for repetition in the data in the block. In some cases, the apparatus is provided with means for counting and encoding the number of repetitions and means for generating a compressed code from these encoded codes. To.

According to a third aspect of the present invention, there is provided an image data compression method for compressing pseudo-tone image data composed of two-dimensionally arranged pixels. Detect the size, determine the block unit to execute encoding from the size of the detected pseudo gradation pattern, divide the compression target image into the block unit, and if the data in this block unit has repetition, It is characterized in that the number of repetitions is counted and encoded, and when there is no repetition, data in block units is encoded.

According to a fourth aspect of the present invention, in the image data compression method according to the third aspect, the size of the block unit is such that the size of the detected pseudo gradation pattern is in the main scanning direction (N1). In the sub-scanning direction (N2), N1 ×
M (M ≦ N2).

According to a fifth aspect of the present invention, in the image data compression method according to the fourth aspect, when the value in the sub-scanning direction (N2) is two, the value of M is set to two, and Is divided into N1 × 2 blocks without duplication.

According to a sixth aspect of the present invention, in the image data compression method of the fourth aspect, when the value in the sub-scanning direction (N2) is three, the value of M is set to three, and Is divided into N1 × 3 blocks without duplication.

According to a seventh aspect of the present invention, in the image data compression method according to the fourth aspect, when the value in the sub-scanning direction (N2) is 4 or more, the value of M is set to 4, and Data is divided into N1 × 4 blocks.

[0016]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention is characterized in that a pseudo gradation pattern and its repetition number are separately encoded. As described above, for the pseudo gradation, a method is used in which a square grid is usually set according to the gradation level of the original image, and the filling method is changed according to the gradation value. When the values are the same, the same pattern is repeated with the size of the pseudo gradation pattern (hereinafter referred to as dither block size) as a cycle. According to the present invention, data comparison is performed in block units corresponding to the pseudo gradation pattern, repetition is detected, and the number of repetitions is encoded.
This method is similar to the technique disclosed in Japanese Patent Application Laid-Open No. 9-65147. However, since the data to be encoded is two-dimensional block data, it is possible to obtain a higher compression ratio than this conventional technique. This is a new technology that becomes possible.

In order to execute the data comparison in block units, it is necessary to know the dither block size.
It is also a feature of the present invention that a means for detecting this is provided. As described above, the gray level of the original image is often set to a power of 2 (16 gray levels / 64 gray levels / 128 gray levels / 256 gray levels), and the dither block size in this case is 4 × 4, 8 × 8, 11 in order
× 11, 16 × 16. (Here 11x
Although 11 = 121, since the value is close to 128, it seems that this dither block size is often used. If this information is not included as data in the compression target image, it is essential to detect it. An example of the detection method will be described in an embodiment described later.

Considering only the improvement of the compression ratio and the simplicity of the algorithm, it is preferable that the block unit for detecting the repetition be equal to the dither block size. Since it is necessary to increase the hardware of the comparing means or the speed of encoding and decoding becomes slow, a block unit used for data holding and comparison (hereinafter referred to as a comparison block) is 4 × It is preferable to make the size smaller than the dither block size except for the case of 4 or less. However, in the main scanning direction, the dither block size and the comparison block size need to match.

A normal dither block is 4 × as described above.
Considering that they are set to 4, 8 × 8, 11 × 11 and 16 × 16, 4 × 4 is 4 × 4, 8 × 8 is 4 × 8, and 16 × 16 is 4 × 4. If it is set to × 16, compression encoding of the target image can be performed without special processing. 11x11
Only in the case of, a special process of compressing and encoding the target image by overlapping the comparison blocks is necessary because 11 is a prime number, but the comparison block in this case is also 4 × 1 according to the other cases.
It may be set to 1 (the processing method will be described in an embodiment described later).

In order to detect the repetition of the pattern,
Two data holding means and one data comparing means are required to hold the data of the comparison block. Also, a means for counting the number of times the pattern is repeated is required. According to the present invention, since the pattern of the comparison block and the number of repetitions of the pattern are encoded by different methods, two types of encoding devices corresponding to the encoding are required.
The simplest way to encode the pattern of the comparison block is
In this method, two-dimensional patterns are rearranged one-dimensionally to execute run-length encoding. In addition, in the case of a dither image, the types of patterns appearing in comparison block units are limited (for example, at most 64 types in the case of 4 × 8), the appearance pattern and its sign are uniquely dictionary. It is also possible to adopt a method of registering and registering a new code only when the same pattern as the dictionary is not found even when searching for the same pattern during encoding. For the number of repetitions, a commonly used Huffmann table can be used.

The configuration and operation of the embodiment of the present invention will be described below in detail with reference to the drawings. <Embodiment 1> FIG. 1 is a block diagram showing the configuration of the dither block size detecting means of the present invention. The dither block size detection means includes image data control means 10, block data storage means 20, held data selection means 30, holding means 1 (40), holding means 2 (50), held data comparison means 60, and data pattern storage means. 70. First, a part of the image data to be compressed is stored in the block data storage 20 through the image data controller 10.

Holding means 1 (40) and holding means 2 (50)
Is a means for temporarily holding data of a maximum of 4 × 4 pixels, and the size of the held data is in a range of 2 × 2, 3 × 3, 4 × 4 pixels according to the request of the held data selecting means 30. It is variable. The held data comparison unit 60 is a storage unit
It is determined whether or not (40) and the data of the holding means 2 (50) match, and a request for held data to be compared next is made if the data does not match. If they do not match, the data pattern is registered in the data pattern storage means 70.

When the dither block size is 2 × 2, 3 × 3,
In the case of 4 × 4 pixels, since the number of data patterns (matching the number of gradations) is limited to 5, 10, and 17, respectively, a part or all of the image data to be compressed is divided into its pixel blocks. Data comparison and data pattern registration are executed. If the number of data patterns matches the above numerical value, it is determined that the dither block size is used. Conversely, when the number of registered data patterns exceeds the above value, it can be determined that the size is not at least the dither block size. For example, when performing data comparison and data pattern registration with 2 × 2 pixels, if 6 patterns are registered as shown in FIG. 2, a 2 × 2 dither pattern is not used. (The gradations are 0, 1, 2, 3,
However, in FIG. 2, there are two types of patterns that represent gradation 2. )

When the dither block size is larger than 4 × 4 pixels, first, a place where the pattern matches in 4 × 4 pixel units is detected. For example, when a 6 × 6 dither block as shown in FIG. 3 is used, 4 × 4 block-unit data is stored in the holding unit 1 (40) and the holding unit 2 (5).
Enter 0). The data of the holding unit 1 (40) is fixed, and data comparison is performed while shifting the data selection position of the holding unit 2 (50) one pixel at a time in the main scanning direction. The same pattern should appear for some time. If the same pattern is found, holding means 1 (40) and holding means 2 (50)
Are moved in the same direction by one pixel vertically, horizontally and vertically. If there is a coincidence at the time of moving, the number of moving pixels is increased and the data comparison in 4 × 4 block units is repeated. The designation of the location of the comparison data and the control of the moving direction and the number of moving pixels are performed by the image data control means 10. The location where the data of the holding means 1 (40) fixed at the beginning is taken in is set to various places, and the above-described scanning is repeatedly executed, so that the dither block size is larger than 4 × 4 pixels. Can be detected.

Considering that the generally used dither block size is at most 16 × 16 as a target for executing the above means for detecting the dither block size, the size of the image to be compressed in the main scanning direction is considered. However, it is considered that there is no problem if it is set to about 64 lines in the sub-scanning direction.

<Second Embodiment> After the detection of the dither block size and the determination of the comparison block, the data is compression-encoded over the entire range of the image to be compressed. FIG. 4 is a block diagram illustrating a functional configuration for performing compression encoding according to the present invention. This configuration has a configuration in which a pattern encoding unit 80, a counting unit 90, a counting and encoding unit 100, and a code stream generating unit 110 are added to the configuration of FIG. 1 for detecting the size of a dither pattern. 40) and the holding means 2 (50) can be shared if block data of a maximum of 4 × 16 pixels can be held.

The case where the dither block size is 8 × 8 pixels (see FIG. 5) will be described as an example. The comparison block size is 4 × 8 pixels. For convenience of explanation, numbers are assigned to blocks in units of comparison blocks. N in B (n, m)
And m represent the position of the block with reference to the upper left of the image to be compressed, and B (0,0) is the first block data input of the image data. In FIG. 5, the hatched blocks are 8 × 8 pixel units [for example, B (2,2) and B (3,2)
), The same pseudo gradient is assumed to be B (2,2), B (2,3), B (2,4), B (2,5), B (2,6) , B (4,3), B
(4,4), B (4,5) and B (3,2), B (3,3), B (3,4), B (3,5), B (3,6), B ( 5,3), B
(5,4) and B (5,5) should have the same pattern.

In FIG. 4, the block data held in the holding means 1 (40) is "BL1",
Assuming that the block data held in 0) is “BL2” and the number of repetitions of the detected block pattern is “Rep”, the compression encoding algorithm may execute the processing according to the flowchart of FIG.

Processing is started from BL1 = B (0,0) and BL2 = B (0,1), and data is compared between adjacent blocks. Unless BL1 = BL2, the encoding of BL1 is executed. When BL1 = BL2, the repetition of the block is detected, and the repetition number “Rep” is counted up until the repetition of the same pattern ends. At the end of the repetition, Rep encoding is performed.

In the case of the image data shown in FIG.
When 1 = B (2,2), BL2 = B (2,3), repetition is detected, and the state of BL1 = BL2 is BL1 = B (2,5), BL
Since 2 = B (2,6), the number of repetitions is Rep = 4. The block data pattern (BL1) may be encoded by converting it into a one-dimensional run length each time, or by encoding and registering a dictionary each time a new pattern appears, and referring to it at any time. Number of repetitions (Rep)
Can use a general Huffmann code, but it is necessary to add special data adjacent to the Rep code word so that the code word of BL1 and the code word of Rep can be distinguished during decoding. The code stream generation unit 110 connects the code word of BL1 and the code word of Rep.

In the flowchart of FIG. 6, n is a numerical value representing a block number in the sub-scanning direction. In this case, when the block in the main scanning direction is completed (End of Line), the repetition number is reset to zero. But
With a slight change, a method of counting up the number of repetitions over the line (n) of the scanning block is also possible.

<Embodiment 3> When the dither block size is a multiple of 4, the image to be compressed is 4 × N (N
Can be divided into the dither blocks without overlapping, but in other cases, it is necessary to partially overlap the blocks to be scanned. For example, 11 × 11
In the case of a pixel dither block, repetition of a dither pattern can be detected by dividing the pixel so as to overlap by one pixel as shown in FIG. At the time of decoding, it is necessary to process in consideration of the data duplication part.

By configuring the embodiment as described above, the following effects can be achieved. (1) By providing a means for detecting the size of the pseudo gradation pattern, it is possible to realize a high compression ratio compression apparatus in which the two-dimensional redundancy of pseudo gradation data is reduced. (2) The detection of the size of the pseudo gradation pattern and the efficient encoding of the image data can be realized with the same configuration. (3) By separately encoding the repetition of the block pattern image data and the block pattern, it is difficult to obtain a high compression ratio by ordinary run-length encoding, so that two-dimensional redundancy in the pseudo gradation image data is difficult. A compression method with a reduced compression rate and a high compression ratio can be realized. (4) An image data compression apparatus capable of high-speed processing with a small hardware scale can be realized by dividing and processing an image to be compressed in block units equal to or smaller than the size of the pseudo gradation pattern. . (5) In the case where the size of the pseudo gradation pattern is 2 × 2 pixels or 3 × 3 pixels, the size of the pseudo gradation pattern is made the same as that of the coding block unit, so that the image data having a simple configuration is obtained. A compression device becomes feasible. (6) When the size of the pseudo gradation pattern is 4 × 4 pixels or more, the size of the coding block unit is set to 4
× K pixels (K is the number of pixels of the pseudo gradation pattern in the main scanning direction)
By doing so, it becomes possible to realize an image data compression device capable of high-speed processing with a small hardware scale.

[0034]

As described above, according to the present invention,
It has become possible to obtain a high compression ratio even in the two-dimensional processing of the pseudo gradation image while taking advantage of the conventional technology of the image data compression apparatus.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a dither block size detection unit according to the present invention.

FIG. 2 is an example of a data pattern of a 2 × 2 pixel block.

FIG. 3 is an example when a dither block of 6 × 6 pixels is used.

FIG. 4 is a block diagram illustrating a configuration of compression encoding according to the present invention.

FIG. 5 is an example of image data used for explaining compression encoding.

FIG. 6 is a flowchart for explaining a compression encoding processing procedure of the present invention.

FIG. 7 is an example of a case where scanning blocks are partially overlapped.

FIG. 8 is a repeated example of a black-and-white pattern for explaining a conventional technique.

FIG. 9 is an example for explaining a conventional technique for reducing redundancy in two dimensions.

[Explanation of symbols]

 Reference Signs List 10 image data control means 20 block data storage means 30 held data selection means 40 holding means 1 50 holding means 2 60 held data comparison means 70 data pattern storage means 80 pattern encoding means 90 counting means 100 count encoding means 110 code stream generation means

Claims (7)

[Claims] 1. An image data compression apparatus for compressing pseudo gradation image data composed of two-dimensionally arranged pixels, comprising means for detecting the size of the pseudo gradation pattern. apparatus. 2. An image data compression apparatus for compressing pseudo gradation image data composed of two-dimensionally arranged pixels.
Means for setting a dimensional block to detect the size of the pseudo gradation pattern from part or all of the data of the image to be compressed, and a block unit for performing encoding based on the detected size of the pseudo gradation pattern. Determining, dividing the compression target image into the determined block units, and determining whether or not the data in the block unit has repetition. When the data in the block unit does not have repetition, the data in the block unit is determined. Encoding means, counting means for counting the number of repetitions, if the block unit data has repetition, and means for generating a compressed code from these encoded codes. Characteristic image data compression device. 3. An image data compression method for compressing pseudo-gradation image data composed of pixels in a two-dimensional array, wherein the size of the pseudo-gradation pattern is detected from part or all of the data of the image to be compressed and detected. Determine the block unit to execute encoding from the size of the pseudo gradation pattern, divide the compression target image into the block unit, and if there is repetition in the data of this block unit, count the number of repetitions An image data compression method characterized by encoding data in a block unit when the data is encoded and there is no repetition. 4. The image data compression method according to claim 3, wherein the size of the block unit is such that the size of the detected pseudo gradation pattern is the main scanning direction (N1) × the sub-scanning direction (N2). In this case, N1 × M (M ≦ N2). 5. The image data compression method according to claim 4, wherein when the value in the sub-scanning direction (N2) is 2, the value of M is set to 2 and the image data to be compressed is divided into N1 × 2 blocks. A method for compressing image data, wherein the image data is divided without duplication. 6. The image data compression method according to claim 4, wherein when the value in the sub-scanning direction (N2) is 3, the value of M is set to 3, and the compression target image data is divided into N1 × 3 blocks. A method for compressing image data, wherein the image data is divided without duplication. 7. The image data compression method according to claim 4, wherein when the value in the sub-scanning direction (N2) is 4 or more,
An image data compression method characterized in that the compression target image data is divided into N1 × 4 blocks by setting the value of M to 4.