JP2012095227A - Image processing system, image processing method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image processing system in which the number of times of memory access is reduced at decompression of the compressed data that has been compressed in multiple stages, with circuit scale being suppressed from growing too large.SOLUTION: The image processing system includes: representative color decoding means which determines a position for decompressing a representative color obtained by N-th encoding based on the arrangement information available by first-Nth encodings, and decompresses the representative color at the determined position; and N pieces of interpolating color decoding means which determines the position for decompressing the interpolation color obtained by Mth (M is any of integers 1-N) encoding based on the arrangement information available by first-Mth encodings, and decompresses the interpolation color obtained by the Mth encoding at the determined position. The representative color decoding means and the N pieces of interpolation color decoding means operate separately from each other.

Description

  The present invention relates to a technique for compressing an image in predetermined block units and a technique for performing image processing and decoding processing on an image compressed in block units.

  Conventionally, there is a high demand for high-resolution color images, and digital multifunction peripherals are increasingly handling images with a resolution of 1200 dpi or higher in order to meet the demand for higher image quality. As the resolution of an image increases, there is a problem that the number of pixels that require image processing increases dramatically and the processing load increases. Image processing devices such as digital cameras and facsimile machines, not limited to digital multifunction devices, compress color image data in order to save memory and hard disk space and shorten the time required for writing to them. Has been realized.

  As a color still image compression method, a JPEG method using discrete cosine transform or a method using Wavelet transform is often used. In this type of encoding method, generally, an image is encoded into a predetermined block (for example, 8 × 8 or 16 × 16 pixel unit), and high compression efficiency is achieved by performing discrete cosine transform, quantization, and entropy encoding. Have achieved. Since this type of encoding method is a variable-length encoding method, the code amount changes for each image to be encoded. When these image compressions are used, in order to refer to the pixel data and convert the pixel data, it is necessary to decode the compressed data. In other words, image processing cannot be performed with compressed data as it is, and decoding processing is necessarily required, and it is necessary to perform processing on a pixel basis for all pixels of high resolution data, which increases processing time. Invite.

  As a technique for performing compression processing, a known run-length compression method for storing pixel data and the number of continuous data, or a technique for compressing by detecting edges in units of blocks and storing two colors of the edges are disclosed. (For example, Patent Document 1 and Patent Document 2).

JP 2008-271046 A JP-A-10-257488

  In Patent Document 1, the inside of a block is made into two colors, and shape information relating to the arrangement of the two colors and color information of the two colors are stored. On the other hand, in order to achieve higher image quality and higher processing speed, Japanese Patent Application No. 2009-212444, which is a prior application by the applicant of the present application, realizes high compression without degrading the image quality of blocks having two or more colors. is doing. In Japanese Patent Application No. 2009-212444, first, image data is divided into blocks (for example, blocks of 2 × 2 pixel size). Then, by comparing the color data of each pixel in the block, the arrangement pattern information of the color data included in the block of interest and the color data information for the number of colors included in the block of interest are output. Among the output color data information, the first color data information corresponding to the pixel at a predetermined specific position of each block and the other color data information (in the case of a 2 × 2 pixel block, the second to second color data information). 4 color data information) and stored in different memory areas. Therefore, arrangement pattern information, first color data information, and other color data information are collectively stored in the memory area. As a result, thinned-out image data (that is, color data information of the first color) at a specific position is placed in the continuous memory area, so that complex decoding processing is required only by reading out the color data information of the first color. Can handle low-resolution images with half the resolution.

  On the other hand, Japanese Patent Application No. 2009-212444 also describes a method of decompressing data compressed by the compression technique. For example, as shown in FIG. 19, first, the arrangement pattern information of the block of interest and the first color data information are read from the memory, and further, the other color data information (the first number of colors) corresponding to the required number of colors based on the arrangement pattern information. 2-4 color data information). Then, in accordance with the arrangement pattern information, the first color data information and the other color data information are arranged on the memory to perform data expansion (decoding).

  The present applicant also considers multi-stage compression by repeatedly applying the compression technique proposed in Japanese Patent Application No. 2009-212444 a plurality of times. The multi-stage compression method will be described in detail in an embodiment to be described later, but will be briefly described as follows. That is, by applying the compression process (first-stage compression process) described above to the original image data, the first-stage arrangement pattern information, the first color data information, and the other color data information And convert to Further, the second-stage compression processing is performed using the color data information of the first color obtained by the first-stage compression processing as the next processing target image. If such multi-stage compression is performed, the color data information of the first color obtained by the compression process of the first stage becomes the arrangement pattern information of the second stage by the compression process of the second stage and the first color of the second stage. Color data information and other color data information in the second stage. For example, when the block size is 2 × 2 pixels, the other color data information is represented by the color data information of the second to fourth colors. Therefore, if the above-described compression process is repeated two stages, FIG. The compressed data as shown in FIG.

  On the other hand, when it is configured to use a single decompression unit to decompress data that has been compressed in multiple stages as described above in a reverse order as compared with the time of compression, the following problems may occur. First, at the time of multi-stage development, the memory bandwidth is compressed by increasing the number of memory accesses. For example, the color data information of the first color after the first stage compression processing for a 16-block processing target image (an image having 4 × 4 blocks of 2 × 2 pixel size) is an image of 4 blocks ( 2 × 2 pixel size image). Further, the color data information of the first color after the second-stage compression processing is performed on the four-block image becomes a one-block image. When the compressed data compressed into one block is developed in two stages and developed into the original image, first, the second stage compressed data is developed (the first color in the second stage according to the arrangement pattern information in the second stage). In this process, at least one block of color data is read and four blocks of color data are written. Further, when the decompression process of the first-stage compressed data is performed with the data as a processing target, at least 4 blocks of color data are read and 16 blocks of color data are written. Therefore, the number of memory accesses increases, and the memory bandwidth is reduced.

  In addition, in order to increase the throughput of multi-stage deployment, in the case where a plurality of development sections described above are provided and pipelined so as to develop in stages, an intermediate arrangement between the first and second development sections As the number of buffers and expansion units increases, the circuit scale increases. For example, one block of information is increased to four blocks by expansion in the expansion unit that expands the second-stage compressed data. Therefore, in order for the decompression unit of the first stage compressed data to perform the decompression process with the same throughput as the decompression unit of the second stage compressed data, four decompression units that decompress the first stage compressed data are required. An intermediate buffer for 4 blocks is also required. Therefore, five expansion units are required as a whole, and the circuit scale increases including the intermediate buffer. Further, when decompressing data compressed in three or more stages, the number of necessary decompression units increases exponentially and the circuit scale increases, which makes it difficult to implement.

  That is, if compressed data that has been subjected to multi-stage compression is expanded in a stepwise manner by one expansion unit, the number of accesses to the main memory increases and the memory bandwidth is compressed. In addition, when a plurality of development units are pipelined and development is performed in stages, the circuit scale, such as an intermediate buffer, increases, which makes implementation difficult.

  In order to solve the above problems, the present invention has the following configuration. That is, the image data is divided into blocks each having a 2 × 2 pixel size, the pixel value of a pixel at a predetermined position in each divided block is a representative color, and the pixel value other than the representative color in each block is an interpolation color. The encoding unit that acquires the arrangement pattern of the representative color and the interpolation color in each block as arrangement information and stores the representative color, the interpolation color, and the arrangement information in different storage areas respectively, An image processing apparatus for decoding compressed data obtained by performing N-th coding on image data composed of representative colors obtained by encoding (N is an integer of 2 or more), wherein the code The position where the representative color obtained by the Nth encoding by the encoding means should be developed is based on the arrangement information obtained by the 1st to Nth encodings by the encoding means. The representative color decoding means for developing the representative color at the specified position and the interpolation color obtained by the Mth encoding (M is an integer from 1 to N) by the encoding means are developed. The position to be determined is specified based on the arrangement information obtained by the 1st to Mth encodings by the encoding means, and the interpolation color obtained by the Mth encoding is developed at the specified position. And N interpolation color decoding means, wherein the representative color decoding means and the N interpolation color decoding means operate independently of each other.

  Provided is a decoding method in which the number of memory accesses is small when multi-stage compressed data is expanded, and the implementation for speeding up is easy without increasing the circuit scale.

1 is a diagram showing an outline of an MFP system according to the present invention. The figure which showed the outline | summary of the controller which concerns on this invention. The figure which enumerated the pattern of the block when dividing an image into blocks. The figure which enumerated the pattern of the block and its identifier. The figure which showed the flow of the image compression which concerns on 1st embodiment. The figure which showed the relationship between the input and output with respect to 1 step | paragraph compression processing which concerns on 1st embodiment. The figure which shows the layout on the memory space of the 1 step | paragraph compression data which concern on 1st embodiment. The figure which shows the layout on the memory space of the 2 step | paragraph compression data which concern on 1st embodiment. The figure which shows the detailed block diagram of the expansion | deployment part 211 which concerns on this invention. The figure which shows the expansion | deployment processing flow of the 1 step | paragraph compression data based on this invention. The figure which shows an example of the layout on the memory space of 1 step | paragraph compression data. The figure which shows the expansion | deployment pixel position output at the time of expansion | deployment of 1 step | paragraph compression data. The figure which shows the expansion | deployment processing flow of the two-step compression data based on this invention. The figure which shows an example of the layout on the memory space of 2 step | paragraph compression data. The figure which shows the expansion | deployment pixel position output at the time of expansion | deployment of 1 step | paragraph compression data. The figure which shows the comparison of the memory access amount by the expansion | deployment part of a conventional structure and this invention. The block diagram in the case of performing in parallel the expansion | deployment part of a conventional structure and this invention. The block diagram at the time of carrying out in parallel the expansion | deployment part which concerns on this invention. The detailed block diagram of the expansion | deployment part of a conventional structure. The figure which shows the expansion | deployment processing flow of the 1 step | paragraph compression data which concern on the expansion | deployment part of a conventional structure. The figure which shows the example of an image of the repeat flag which concerns on 2nd embodiment. The figure which showed an example of the layout on the memory space of the two-step compression data at the time of applying the repeat flag which concerns on 2nd embodiment.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. In addition, the structure shown below is an example for implement | achieving this invention, and a structure is not limited to this. In the present embodiment, a digital multifunction peripheral (MFP) having a plurality of functions such as scanning, printing, and copying will be described as an example of the image processing apparatus.

[First embodiment]
As shown in FIG. 1, the controller 101 is connected to a scanner 102 as an image input device and a printer 103 as an image output device. Further, the controller 101 is connected to a network 104 such as a LAN or a public line (WAN) to input / output image information and device information and develop an image of PDL data. The CPU 105 is a processor that controls the entire MFP by executing a program stored in an HDD storage unit 107 described later. The memory 106 is a system work memory for operating the CPU 105, and is also an image memory for temporarily storing image data. The HDD storage unit 107 is a hard disk drive, and stores system software, programs, image data, and the like.

  Next, detailed processing of each unit of the controller 101 will be described with reference to a configuration example of the controller 101 shown in FIG. First, a case where image data scanned by the scanner 102 is read will be described. The scanner image processing unit 201 receives RGB (red, green, blue) three-color image data (raster image) read by the scanner 102. Then, image processing such as shading processing and filter processing is performed on the image data, and the compression unit 202 performs compression (encoding) processing. The data (compressed data) compressed by the compression unit 202 is stored in the memory 106 via the image memory bus.

  Next, when printing the scanned image data, the compressed data stored in the memory 106 is transferred to the expansion unit 211 via the image memory bus. The expansion unit 211 expands (decodes) the data compressed by the compression unit 202 and transfers the data to the color processing unit 212. Then, the color processing unit 212 converts into a CMYK (cyan, magenta, yellow, black) color space. Thereafter, the color processing unit 212 further performs color processing such as density adjustment and printer gamma correction on each value of CMYK, and then transfers the data after the color processing to the compression unit 202 to perform compression processing again. Then, it is stored again in the memory 106 via the image memory bus. Thereafter, in order to perform image processing for printing, raster image data stored in the memory 106 is transferred via the image memory bus to the developing unit 211 and the printing image processing unit 221 in this order, and the developing processing and printing Perform image processing. The print image processing unit 221 performs area gradation processing by the dither method or the error diffusion method on the input raster CMYK image data, and outputs the processed data to the printer 103. The area gradation processing here is not limited to the method described above, and area gradation processing by another method may be applied as long as the present invention is applicable.

  When the scanned image data is transmitted to the network, the compressed data stored in the memory 106 is transferred to the expansion unit 211 via the image memory bus. The developed raster data is transferred to the color processing unit 212. The color processing unit 212 performs display gamma adjustment, paper background color adjustment, and the like, and then converts to a YCbCr (luminance, BLUE color difference, RED color difference) color space. Then, the data processed by the color processing unit 212 is transferred to the compression unit 202, compressed again, and stored again in the memory 106 via the image memory bus. Thereafter, the compressed data stored in the memory 106 is transferred to the expansion unit 211 via the image memory bus in order to perform image processing for transmission. Then, the expansion unit 211 expands the compressed data into raster image data. Thereafter, if the transmission processing unit 233 performs color image transmission on the raster YCbCr image data, JPEG compression processing is performed. If the monochrome binary image transmission is performed, binarization is performed on the Y data to perform JBIG compression or the like. And output to the network 104.

  When the scanned image data is saved, the compressed data stored in the memory 106 is transferred to the disk spool high compression / decompression unit 242 via the image memory bus. In the disk spool high compression / decompression unit 242, since the writing speed to the HDD is slower than the writing speed to the memory, a higher compression JPEG compression is performed. Thereafter, the compressed data is stored in the HDD storage unit 107 via the disk access controller 243. In addition, when the compressed data stored in the HDD storage unit 107 is transferred again to the memory 106, the above-described processing may be performed in reverse.

  Here, a case where PDL data sent from another device connected via the network 104 shown in FIG. Although the PDL interpretation unit is not shown in FIG. 2, the CPU 105 functioning as the PDL interpretation unit interprets the PDL data and outputs the display list as a result to the memory 106. Thereafter, the rendering unit 253 renders the display list stored in the memory 106 as raster RGB image data, and the compression unit 202 performs image compression processing on the rendered RGB image data (raster image data). . Then, the compressed data after the image compression processing is stored in the memory 106 via the image memory bus. It should be noted that the process of printing, transmitting to a network, and storing the PDL data can be realized by performing the same process as in the case of scanned image data.

[Compression processing]
Next, compression processing by encoding raster image data will be described in detail. In the present embodiment, all compression processing is performed by the compression unit 202, but the present invention is not limited to this configuration. For example, instead of performing all the compression processing in the compression unit 202 as shown in FIG. 2, a plurality of compression processing units may be provided so that the compression units are distributed and used for each image path. . In this embodiment, first, raster image data in units of pages is divided into blocks of 2 × 2 pixels, and data compression processing is performed in units of blocks extracted by division.

  Here, before describing the compression processing in this embodiment, the number of combinations is considered according to the number of colors occupied in a block of 2 × 2 pixels. Here, since the number of pixels included in the block is four, the maximum number of colors in each block is four, and there are at most combinations of one to four colors in the block. FIG. 3 is a diagram illustrating an arrangement pattern of colors in a block that can occur when the number of colors is 1 to 4 respectively.

  First, when the number of colors in a block is one, all four pixels are composed of the same color, so there is only one combination.

  Next, consider a case where the number of colors in a block is two (each color will be referred to as a first color and a second color). Considering the color data of a pixel at a predetermined position in the block (in this embodiment, the upper left pixel in the block) as the first color, the first color or the remaining three pixels other than the upper left pixel are changed to the first color or Since the second color is included, there are seven possible combinations except for the case of four pixels having the same color.

  Next, consider a case where the number of colors in a block is three (first to third colors). The combination in the case where three colors are laid out in four pixels can be rephrased as the number when only one of the three colors is used twice. Of the coordinates of the four pixels, the two pixels have the same color. What is necessary is just to obtain | require the number of cases. That is, the number of combinations in the case of three colors is a combination that takes two coordinates from four coordinates, and there are six combinations in total. Further, there are only one combination when the number of colors in the block is four (first to fourth colors).

  As described above, when the number of combinations in all cases where the number of colors in a block of 2 × 2 pixels is 1 to 4 is added up, 15 arrangement patterns can be considered in total. Further, considering that a flag (identifier) is added to identify all these patterns, 4 bits are required as the data amount of the flag. The relationship between these 15 patterns and flags is defined as shown in FIG. 4, and a flag indicating the arrangement information is hereinafter referred to as a “pattern flag”. In addition, the first color is collectively referred to as “representative color”, the second color, the third color, and the fourth color (that is, colors other than the representative color) in the block. Note that the number of pixels constituting the block is not limited to 2 × 2 pixels, and may be larger than this size. In this case, since the number of arrangement pattern combinations increases, the number of bits of the pattern flag expressed by 4 bits also increases.

[Compression processing flow]
As described above, the processing performed by the compression unit 202 will be described using the compression processing flow (FIG. 5) of the compression unit 202 based on the combinations (color arrangement patterns) that can be taken in the 2 × 2 pixel block. . The input in the present embodiment will be described as image data having, for example, 256 gradations of 8 bits for each of RGB and image data having a color value (pixel value) of 24 bits per pixel in a dot sequential manner of 8 bit data. In this embodiment, the compression unit 202 exists in the controller 101 as shown in FIG. 2 and is configured by hardware (such as an ASIC) such as an electronic circuit. However, as another configuration, a part or all of the functions may be realized by the CPU executing the program.

  First, this processing is started, and an image divided into blocks each having a 2 × 2 pixel size is input from the scanner image processing unit 201 or the rendering unit 253 in the previous stage (S501). Then, the compression unit 202 performs color reduction processing and pattern flag calculation processing in the 2 × 2 pixel block (S502). The color reduction processing in S502 is performed by changing the compression efficiency of the main compression processing by changing them to the same color (pixel value) when there are a plurality of pixels having similar pixel values in 2 × 2 pixels. To increase. The color reduction processing here is not particularly limited, and any method may be used as long as the present invention is applicable. Then, by comparing the color data of the block composed of 2 × 2 pixels after the color reduction processing, a pattern flag indicating the number of colors after color reduction and the color arrangement pattern in the block is calculated.

  Next, based on the pattern flag calculated in S502, the compression unit 202 determines the number of colors and the arrangement pattern in the block, and extracts color data of each color (S503). As shown in FIG. 4, since the number of colors constituting the block and the arrangement pattern of each color are associated with the 4-bit pattern flag, the number of colors in each block and the color of each color from the pattern flag. Data can be identified.

  Here, in FIG. 4 of the present embodiment, it is defined that the color (pixel value) of the upper left pixel is the first color (first color data) in all patterns. When the pattern flag is “0”, since the number of colors is 1, the color of the upper left pixel (pixel value) is extracted as the first color. When the pattern flag is 1 to 7, since the number of colors is 2, the color of the upper left pixel (pixel value) is extracted as the first color, and further, the second color defined according to each pattern flag The color (pixel value) of the pixel at the position where (second color data) exists is extracted. For example, when the pattern flag is “1”, the color of the upper right pixel in the block is extracted as the second color (second color data).

  When the pattern flag is 8 to D, since the number of colors is 3, the color (pixel value) of the upper left pixel in the block is extracted as the first color. Further, the color (pixel value) of the pixel at the position where the second color (second color data) and the third color (third color data) defined according to each pattern flag exist is extracted. For example, when the pattern flag is “8”, the color of the upper left pixel in the block is the first color, the color of the upper right pixel (pixel value) is the second color, and the color of the lower right pixel (pixel value) is the first color. Extract as 3 colors.

  When the pattern flag is “E”, since the number of colors is 4, the color of the upper left pixel in the block is the first color, the color of the upper right pixel is the second color, and the color of the lower left pixel is the third color. The color and the color of the lower right pixel are extracted as the fourth color.

  In other words, based on the pattern flag, the compression unit 202 specifies the number of colors constituting the block (S504, S506, S508), and outputs a pattern flag and color data according to the specified number of colors. (S505, S507, S509, S510). Then, this processing flow ends.

  The output data will be described with reference to FIG. Here, each of the 2 × 2 pixels has a pixel value of 24 bits (8 bits for each color of RGB), and the input data has a data amount of 96 bits per block. On the other hand, as shown in FIG. 6, for example, when the pattern flag is “0” (that is, four pixels of the block are configured with one color) (YES in S504 in FIG. 5), the color data for the second and subsequent colors Does not exist. Accordingly, 4 bits of the pattern flag and the pixel value of the first color (24-bit color data) are output (S505 in FIG. 5). That is, the block having the pattern flag “0” outputs data having a data amount of 28 bits. When the pattern flag is 1 to 7 (that is, four pixels are composed of two colors) (YES in S506 in FIG. 5), the coordinates of the pixel in which the second color exists is obtained from the pattern flag, and the pattern flag 4 Bits and pixel values for two colors (color data for 48 bits) are output (S507 in FIG. 5). That is, the blocks having the pattern flags 1 to 7 output data having a data amount of 52 bits.

  If the pattern flag is 8 to D (that is, three colors) (YES in S508 in FIG. 5), the position of the pixel where the first color, the second color, and the third color exist is obtained based on the pattern flag, and the pattern flag The pixel values for 4 bits and 3 colors (color data for 72 bits) are output (S509 in FIG. 5). That is, a block having a pattern flag of 8 to D outputs data having a data amount of 76 bits.

  When the pattern flag is “E” (that is, four colors) (NO in S508 in FIG. 5), 4 bits of the pattern flag and pixel values for four colors (96-bit color data) are output ( (S510 in FIG. 5). That is, the block having the pattern flag “E” outputs data having a data amount of 100 bits.

  In other words, the color data output from each block appears until then when scanned in the order of position within the block (from left to top, top left, bottom left, bottom right, 1, 2, 3, 4). This is equivalent to outputting the data every time color data that has not been displayed appears.

  As described above, the color data (96 bits) for the four colors of the block composed of 2 × 2 pixels is output as the data (color data) of the pixel values for the number of colors existing in the block and the 4-bit pattern flag. To do. As a result, the amount of output data can be reduced by simple processing. In particular, in the case of a raster image having a large number of blocks having the same color pixel (that is, a block having a small number of colors in each block) in a 2 × 2 pixel block, the compression rate of the output data amount is also increased. Further, by referring to the pattern flag, the number of colors in the block and the arrangement pattern can be specified. By performing such processing on all the image blocks, the entire image can be compressed.

[Storage method]
The pattern flag and the color data obtained and output by the compression unit 202 by the method described above are written in the memory 106 (memory storing process). At this time, the compression unit 202 changes the writing position of the pattern flag, the first color data, and the second color, third color, and fourth color data in the memory 106. The compression unit 202 has three addresses: a memory start address for writing a pattern flag, a memory start address for writing first color data, and a memory start address for writing second to fourth color data. Is specified. That is, the pattern flags of each block are collectively stored in a pattern flag storage unit (memory area for storing pattern flags) on the memory 106. The color data of the first color (first color) of each block is stored together in a first color storage unit (memory area for storing the first color data of each block) on the memory 106. become. Further, the color data of the second to fourth colors of each block are collected in a second, third, and fourth color storage unit (memory area for storing the second to fourth color data of each block) on the memory 106. Will be stored.

  FIG. 7 is a diagram illustrating an example of writing the image data compressed by the compression unit 202 into the memory space. When an 8-bit RGB image having a size of M × N pixels is input to the compression unit for each 2 × 2 pixel block, the data size of the pattern flag storage unit in which the pattern flag data is stored is (M × N / 2/2 × 4/8) bytes. Further, the data size of the first color storage unit in which the first color data is stored is (M × N / 2/2 × 24/8) bytes. The data sizes of the second, third, and fourth color storage units that store the second, third, and fourth color data differ depending on the raster image to be processed. This is because the number of blocks in which the second, third, and fourth colors are present differs depending on the image. That is, since the number of colors included in each block differs depending on the pattern, the number of the second, third, and fourth colors varies, although the pattern flag and the first color are always specified as one with respect to the number of blocks. . Therefore, the size of the output data compressed by the compression unit 202 depends on the amount of data stored in the second, third, and fourth color storage units on the memory 106. In addition, in the memory area (first color storage unit and second, third, fourth color storage unit) after the first color write start address, pixel color data is stored in the same number of bits. Therefore, for example, when performing color processing that is completed with one pixel input and one pixel output, such as color conversion using a lookup table (LUT), gamma correction processing, and color space conversion processing using matrix operation, the compressed data is decoded. The corresponding color data in the compressed data can be processed without (decoding) and returning to the original raster image. That is, as shown in FIG. 7, the corresponding pixel value (color data) portion can be directly processed with respect to the data stored in the first color storage unit and the second to fourth color storage units of the memory. is there.

  Further, as shown in FIG. 7, the image data 700 is divided into a pattern flag, a first color, and other colors (second, third, and fourth colors), and discretely stored in a storage area on the memory 106. Yes. Therefore, in the first color storage unit, pixel values (color data) when the raster image is divided into blocks of 2 × 2 pixels and the upper left pixel of each block is sampled are continuously present in the memory. Will do.

[Multi-stage compression]
Next, a case where multistage compression is performed by the above-described compression method will be described. The multistage compression here means that the same image data is subjected to compression processing a plurality of times (N times: N is an integer of 2 or more). For example, in the two-stage compression, after the first compression process is performed on the input image data, the generated image data of only the representative colors (the first color stored in the first color storage unit) The second compression processing is performed on the next compression processing target image. Therefore, in the N-th compression process, the compression process is performed using image data composed of representative colors obtained by the N-1th compression process. According to the above-described compression method, the first color storage unit after the first compression processing stores the color data of the pixels of the portion shown in white in FIG. 7 in raster order regardless of the compression rate. Yes. That is, data obtained by thinning the original image data every two horizontal and vertical pixels is stored. In the multi-stage compression, the image data generated by the thinning is further divided into blocks of 2 × 2 pixels, and the above-described compression process is applied again.

  That is, as shown in FIG. 8A, attention is focused on data stored in the first color storage unit among the image data compressed by the first stage compression processing. The color data stored in the first color storage unit of the memory after the first-stage compression processing in FIG. 8A is pixel data at positions indicated by white and black in the image data 800. Then, the image data (first color data after the first stage compression process) stored in the first color storage unit is further blocked every 2 × 2 pixels, and the above-described compression process is performed. As a result, the color data of the first color as a result of the compression process of the first stage includes the pattern flag of the result of the compression process of the second stage, the color data of the first color of the second stage, the second color of the second stage. It is compressed and converted into color data of 3 or 4 colors. Therefore, a compression result as shown in the second-stage compression memory space in FIG. 8B is obtained. Here, as shown in FIG. 8B, in the memory space, the color data at the position indicated by white in the image data 800 is stored in the first color storage unit at the time of the second stage compression. In the second, third, and fourth color storage units in the second stage compression, color data other than the first color obtained based on the pixel at the position indicated by black in the image data 800 is stored. However, the second, third, and fourth color storage units in the second stage compression do not store a plurality of the same color data existing in 2 × 2 pixels. The multi-stage compressed data is stored in a nested structure as shown in FIG. 8B. As another storage method, the second-stage compressed data may be stored by newly securing another memory area instead of the same memory area.

[Deployment processing]
Next, the developing unit 211 that forms the core of the present invention will be described. The expansion unit 211 performs processing to return the pattern flag and color data generated by the compression processing as described above to the original raster image data. FIG. 9 shows a detailed block diagram of the expansion unit 211 in the present invention. The expansion unit 211 in the present embodiment exists in the controller 101 as shown in FIG. 2 and is configured by three data expansion units 900, 910, and 920. Note that the expansion unit 211 of this embodiment is configured by hardware such as an electronic circuit (ASIC or the like), but the CPU executes a program to realize part or all of the functions. It doesn't matter.

  First, in the case where compressed data that has been subjected to one-stage compression processing is a decompression target, detailed processing of the decompression unit 211 will be described using the flowchart shown in FIG. 10 and the compressed image example shown in FIG.

  In the present embodiment, in order to simplify the description, a case where the compressed data is expanded will be described by taking compressed data of 4 × 4 pixels as shown in FIG. 11A as an example. FIG. 11B shows an extracted part of the memory map for storing compressed data in the memory 106, and FIG. 11C shows a detailed data area in the memory. In the image data 1101, pixels represented by the same pattern (shaded part, white part, vertical line part, grid line part) indicate pixels having the same pixel value.

  When one-stage compression processing is applied to the image data 1101 shown in FIG. 11A, the image data 1101 is divided and compressed as a thick black line in blocks of 2 × 2 pixels. Then, the pattern flag, the first color data, and the second to fourth color data of each block as a result of compression are stored in the pattern flag storage unit 1102 on the memory 106 shown in FIG. Are stored in the unit 1103 and the second, third, and fourth color storage units 1104, respectively. The pattern flag of each block of the image data 1101 is “B”, “5”, “1”, “E” in the order of the upper left block, the upper right block, the lower left block, and the lower right block according to the assignment shown in FIG. Become. These values are stored in the pattern flag storage unit 1102 and arranged as addresses 0x00 and 0x01 in FIG.

  The first color storage unit 1103 stores pixel value data at the upper left of each block continuously. Accordingly, the pixel values of the pixels 00, 02, 20, and 22 are stored as in the address 0x10-0x1B in FIG. However, in this embodiment, an 8-bit RGB image is handled, and the size of one pixel data is 24 bits (3 bytes). In addition, the data amount of each address shown in FIG. 11C is 8 bits (1 byte).

  In FIG. 11C, the horizontal value indicates the value of the first significant memory address, and the vertical value indicates the value of the second significant memory address. That is, “0x10”, which is the first address of the first color storage unit 1103 shown in FIG. 11B, is “00” in the horizontal direction and “10” in the vertical direction in FIG. It means the address that intersects.

  Finally, the second, third and fourth color storage unit 1104 stores pixel values (color data of a color different from the first color) other than the upper left pixel of each block. However, if the same color exists in the same block, it is not stored redundantly. For example, since the pixel 01 and the pixel 10 in the upper left block of the image data 1101 illustrated in FIG. 11A have the same color, only the pixel values of the pixel 01 and the pixel 11 (that is, color data of a color different from the first color) are the second. , 3 and 4 color storage units. Similarly, since the pixel 02, the pixel 03, and the pixel 13 in the upper right block of the image data 1101 have the same color (first color), only the pixel value of the pixel 12 is stored in the second, third, and fourth color storage units. Stored. Therefore, in the example of the image data 1101 in FIG. 11, the pixel values stored in the second, third, and fourth color storage units are the pixel values (color data) of the pixels 01, 11, 12, 21, 23, 32, and 33. Therefore, it is stored as addresses 0x40-0x54 in FIG.

  In this embodiment, the decompression processing of compressed data that has been subjected to one-stage compression processing as shown in FIG. 11 is performed using two of the three data decompression units 900, 910, and 920 in FIG. Is called. Here, it is assumed that the data expansion unit 900 expands the first color data, and the data expansion unit 910 expands the second, third, and fourth color data. That is, the data expansion unit 900 specifies only pixels that are filled with the first color data in the compressed data based on the pattern flag in the raster image to be expanded. Then, the specified pixel is filled with the read first color data. In addition, the data development unit 910 specifies only pixels that are filled with the second, third, and fourth color data in the compressed data based on the pattern flag in the raster image to be developed. Then, a process for painting each of the specified pixels is performed with the read second, third, and fourth color data. Here, the painting process means a process of applying the value of the color data to the value of each pixel in the development area (development memory area) of the image data.

  First, the operation of the data expansion unit 900 that expands the first color data will be described with reference to FIG. Hereinafter, a block that the data output units 904, 914, and 924 focus on in one expansion process is referred to as a target block. In the decompression processing of compressed data that has been subjected to one-stage compression processing, the target block is a 2 × 2 pixel unit block of the image data 1101.

  The pattern flag DMAC 901 is set with the head address and image size of the pattern flag storage unit 1102, reads the pattern flag in order from the set address (for example, 0x00), and sequentially transfers it to the data output unit 904. (S1001). In parallel, the data DMAC 903 is set with the first address of the first color storage unit 1103 and the image size, reads the first color data sequentially from the set address (for example, 0x10), and sequentially outputs it to the data output unit 904. Transfer processing is performed (S1002). Based on the pattern flag acquired from the pattern flag storage unit 1102, the data output unit 904 specifies the pixel position to be filled with the first color data acquired simultaneously (S1003).

  For example, for the pattern flag “B” first read from the pattern flag storage unit 1102, the data output unit 904 selects only the pixel 00 of the upper left pixel from the correspondence table shown in FIG. 4 as the first color of the block. Is determined to be filled with color data. For the pattern flag “5” read next, the upper left pixel, the upper right pixel, the lower right pixel 02, the pixel 03, and the pixel 13 are similarly assigned to the block number of the block from the correspondence table shown in FIG. It is determined to fill with one color data.

  Next, when there is a pixel to be filled in the target block (Yes in S1004), the data output unit 904 stores the rasterized image start address stored in advance in the data output unit 904 and the size information of the expanded image Then, a memory address corresponding to a pixel to be painted with the first color data is calculated (S1005). Then, the first color data received from the data DMAC 903 is written in the memory area of the calculated address (S1006).

  The processing of S1004 to S1006 is repeatedly executed for the number of pixels to which the first color data is written in one block. That is, in the upper left block of the image data 1101, the pixels for writing the first color data are for one pixel, so the processing of S1004 to S1006 is performed once. Further, in the upper right block of the image data 1101, the pixels for writing the first color data are for three pixels, so the processing of S1004 to S1006 is performed three times. When the processing for one block is completed in this way, in S1004, the data output unit 904 determines that there are no pixels to be filled (No in S1004), and completes the processing for developing the target block.

  With the above processing, the pixels in which the color data of the first color of each block is written as pixel values on the development memory are represented by the hatched portion in FIG. In other words, the pixel value of the pixel represented by the hatched portion in FIG.

  Next, the operation of the data expansion unit 910 that expands the second, third, and fourth color data will be described with reference to FIG. The pattern flag DMAC 911 sets the head address and image size of the pattern flag storage unit 1102, reads the pattern flag in order from the set address (for example, 0x00), and sequentially transfers the data to the data output unit 914 ( S1021). In parallel, the data DMAC 913 is set with the start address and image size of the second, third, and fourth color storage units 1104, and reads the second, third, and fourth color data in order from the set address 0x40. A process of sequentially transferring to the output unit 914 is performed (S1022). Based on the pattern flag of the target block acquired from the pattern flag storage unit 1102, the data output unit 914 first determines whether there is second, third, and fourth color data in the target block (S 1023).

  For example, when there is a block having the pattern flag “0”, since the target block is a block composed only of the first color data, the second, third and fourth color data do not exist in the target block (S1023). No). For this reason, the data development unit 910 does not perform pixel filling processing using the color data of the second, third, and fourth colors of the block of interest, and moves to processing of the next block. If any other pattern flag is read, it is determined that the block has at least the second color data (Yes in S1023).

  Next, in a case where at least the second color data exists (Yes in S1023) and there are still pixels to be filled in the target block (Yes in S1024), the data output unit 914 reads the second and third data read. , The pixel position to be filled with the four color data is specified based on the pattern flag (S1025). Since the data DMAC sequentially transfers the second, third, and fourth color data in one block to the data output unit 914, the data output unit 914 includes pixels that are filled with the second color data, pixels that are filled with the third color data, The pixels to be filled with the fourth color data are specified separately. Further, the data output unit 914 applies the color data of the second, third, and fourth colors to the pixels that are filled with the color data of the second, third, and fourth colors from the start address for storing the rasterized image set in the data output unit 914 and the size information of the developed image. The corresponding memory address is calculated (S1026).

  Then, the data output unit 914 writes the second, third, and fourth color data received from the data DMAC 913 in the memory area of each calculated address (S1027). The processing of S1024 to S1026 is repeated for the number of pixels to which the second, third, and fourth color data are written in one block. That is, in the upper left block of the image data 1101, since the second, third, and fourth color data to be written are for three pixels, the processing of S1024 to S1026 is performed three times. In the upper right block of the image data 1101, the second, third, and fourth color data to be written is for one pixel, so the processing of S1024 to S1026 is performed once. When the processing for one block is completed in this way, in S1024, the data output unit 914 determines that there are no pixels to be filled (No in S1024), and ends the development processing of the target block. Through the above processing, the pixels in which the color data of the second, third, and fourth colors of each block are written as pixel values on the development memory are indicated by hatched portions in FIG. In other words, the pixel data of the pixel indicated by the hatched portion in FIG. 12B is expanded by the data expansion unit 910.

  The data compressed by one stage is expanded by the expansion unit 211 by the method as described above. Each DMAC is used to increase the efficiency of data reading, and a configuration in which each DMAC is deleted and each data output unit sequentially accesses necessary data may be employed. Absent.

[Expansion processing for two-stage compressed data]
Next, in the case where the compressed data that has been subjected to the two-stage compression processing is a decompression target, the detailed processing of the decompression unit 211 will be described using the flowchart shown in FIG. 13 and the compressed image example shown in FIG. . That is, a case will be described in which the two-stage compressed data (FIG. 14B) obtained by compressing the image data 1101 composed of 4 × 4 pixels shown in FIG.

  When the image data 1101 is compressed in two stages, it is stored in the memory 106 as compressed data as shown in FIGS. 14B and 14C. First, a mechanism for generating the decompressed two-stage compressed data will be described.

  FIG. 14B shows an extracted part of the memory map for storing compressed data in the memory 106, and FIG. 14C shows a detailed data area inside the memory. In FIG. 14C, the horizontal value indicates the value of the first-order memory address, and the vertical value indicates the value of the second-order memory address. That is, “0x18”, which is the start address of the second stage first color storage unit 1404 shown in FIG. 14B, is “08” in the horizontal direction and “10” in the vertical direction in FIG. "Means the address where"

  The data stored in the first stage pattern flag storage unit 1402, the first stage second, third, and fourth color storage units 1406 and the generation method thereof are the same as the pattern flag storage unit 1102 in FIG. , 3 and 4 color storage unit 1104 and the same generation method. Therefore, explanation is omitted. As the first color data stored in the first color storage unit 1103 by the first-stage compression processing, the upper left pixel of each block of the image data 1101 is stored. That is, it is synonymous with storing image data 1401 (FIG. 14A) thinned out every two pixels in each of the main scanning direction and the sub-scanning direction.

  Therefore, the second stage compression performs a compression process on the image data 1401 obtained by thinning out by the first stage compression. At this time, the pattern flag of the block of the image data 1401 becomes “2” by the assignment shown in FIG. This is stored in the second stage pattern flag storage unit 1403, and is arranged as address 0x10 in FIG. 14 (c). The second-stage first color storage unit 1404 stores the pixel value (color data) of the upper left pixel of the processing target block for the second-stage compression. Accordingly, the pixel data of the pixel 00 is stored like the address 0x18-0x1A in FIG.

  Finally, the second stage second, third, and fourth color storage unit 1405 stores color data (second to fourth colors) other than the color data (first color) of the upper left pixel in the target block to be processed in the second stage compression. ) Is stored. However, if the same color exists in the same block, the data of that color is not stored redundantly. For example, the pixel 00 and the pixel 22 of the block of the image data 1401 have the same color value (first color data), and the pixel 02 and the pixel 20 have the same color value (second color data). Therefore, only the pixel value (second color data) of the pixel 02 is stored in the second stage second, third, fourth color storage unit 1405. Accordingly, the pixel data stored in the second stage second, third, and fourth color storage unit 1405 in the image data 1401 is only the pixel 02, and is stored as the address 0x20-0x22 in FIG. Has been.

  The decompression processing of the compressed data (FIGS. 14B and 14C) subjected to the two-stage compression processing is performed using all the three data decompression units 900, 910, and 920 shown in FIG. Here, the data development unit 900 develops the second stage first color data, the data development unit 910 develops the second stage second, third, and fourth color data, and the data development unit 920 performs the first stage first color data. It is assumed that 2, 3 and 4 color data are developed. That is, the data development unit 900 uses the second-stage pattern flag and the first-stage pattern flag to represent pixels that are filled in the data size after development by the second-stage first color data in the compressed data in the raster image after development. Identify based on. Then, the specified pixel is filled with the read second-stage first color data. In addition, the data expansion unit 910 uses the second-stage pattern flag and the first-stage pattern flag to indicate pixels that are filled with the second-stage second, third, and fourth color data in the compressed data in the raster image after development. Identify. And the process which fills the said specified pixel with the read 2nd step 2nd, 3rd, 4th color data is performed. Further, the data expansion unit 920 identifies pixels to be filled with the first-stage second, third, and fourth color data in the compressed data in the raster image after development based on the first-stage pattern flag. And the process which fills the said specified pixel with the read 1st step 2nd, 3rd, 4th color data is performed.

  First, the operation of the data expansion unit 900 that expands the second stage first color data will be described with reference to FIG. Here, unlike the compressed data decompression process in which the one-stage compression process is performed, in the decompression process of the compressed data in which the two-stage compression process is performed, the target block to be noted in one decompression process Corresponds to a 4 × 4 pixel unit in the image data 1101. In the following description, the processing of S1301 to S1303 is described as being performed in parallel, but the processing can also be performed in order as shown in FIG.

  The pattern flag DMAC 901 is set with the head address and image size of the first stage pattern flag storage unit 1402, reads the pattern flag from the set address (0x00), and sequentially transfers it to the data output unit 904 (S1301). In parallel, the pattern flag DMAC 902 is set with the head address and image size of the second stage pattern flag storage unit 1403, reads the pattern flag from the set address (0x10), and sequentially transfers it to the data output unit 904 ( S1302).

  In parallel, the data DMAC 903 is set with the head address and image size of the second stage first color storage unit 1404, and reads the second stage first color data in order from the set address (0x18). The data is sequentially transferred to the output unit 904 (S1303).

  The data output unit 904 specifies a pixel position to be filled with the second-stage first color data in the developed 4 × 4 pixel size image data based on the acquired first-stage pattern flag and second-stage pattern flag. (S1304). Thereby, a 1st specific means is implement | achieved. Specifically, in the example of the compressed data in FIG. 14C, first, the second stage first color among the pixels 00, 02, 20, and 22 based on the second stage pattern flag (“2”). Pixels 00 and 22 can be specified as pixel positions to be filled with data. Further, the data output unit 904, based on the first stage pattern flag, for the block of the first stage compression including the pixel position specified to be filled with the second stage first color data based on the second stage pattern flag. The pixel position to be filled with the second stage first color data is specified. In the example of FIG. 14C, the first stage pattern flag in the first stage compression block including the pixel 00 is “B”, and the first stage in the first stage compression block including the pixel 22 is set. The pattern flag is “E”. Therefore, the pixels to be filled with the second-stage first color data for the pixel positions other than the pixels 00 and 22 of the block including the pixels 00 and 22 (the blocks corresponding to the upper left block and the lower right block of the image data 1101) are Judge that it does not exist. Therefore, in the example of the compressed data in FIG. 14C, it is possible to specify that only the pixels 00 and 22 are filled in with the second-stage first color data in the expanded 4 × 4 pixel size image data.

  Next, if there is a pixel that has not yet been subjected to the second-stage first-color data filling process (writing process) among the pixels specified in S1304 (Yes in S1305), one of the unprocessed pixels. The process proceeds to S1306. The data output unit 904 is a second-stage first color data from the start address for storing the rasterized image set in advance in the data output unit 904, the size information of the developed image, and the pixel position of the pixel to be processed. A memory address corresponding to the pixel to be painted is calculated (S1306). This implements a first memory address calculation means.

  Then, the data output unit 904 writes the second-stage first color data received from the data DMAC 903 in the memory area of the calculated address (S1307). Thereby, the first storage means is realized.

  The processing of S1306 to S1307 is repeated for the number of pixels to which the second stage first color data is written in one block. In the example of FIG. 14C, in the expanded 4 × 4 pixel size image data, the second-stage first color data is written for two pixels 00 and 22, so the processing of S 1306 to S 1307 is 2. Performed once.

  As described above, when the processing for one target block (corresponding to 4 × 4 pixels because of two-stage compression) is completed, the data output unit 904 fills the second-stage first color data in S1305. It is determined that there are no pixels (the painting process has been completed) (No in S1305), and the process for the block of interest is completed. With the above processing, in the case of the compressed data in FIG. 14C, in the image data expanded on the memory, the pixel to which the second stage first color data is written is represented by the hatched portion in FIG. The In other words, the pixel data of the pixel represented by the hatched portion in FIG.

  Next, the operation of the data expansion unit 910 that expands the second stage second, third, and fourth color data will be described with reference to FIG. In the following description, the processing of S1321 to S1323 is described as being performed in parallel, but the processing can also be performed in order as shown in FIG. 13B.

  The pattern flag DMAC 911 is set with the head address and image size of the first stage pattern flag storage unit 1402, reads the pattern flag from the set address (0x00), and sequentially transfers it to the data output unit 914 (S1321). In parallel with this processing, the pattern flag DMAC 912 is set with the head address and image size of the second stage pattern flag storage unit 1403, reads the pattern flag from the set address (0x10), and sends it to the data output unit 914. Transfer sequentially (S1322).

  In parallel, in the data DMAC 913, the start address and image size of the second stage second, third, and fourth color storage units 1405 are set, and the second stage second, third, and fourth are set from the set address (0x20). The color data is read and sequentially transferred to the data output unit 914 (S1323).

  The data output unit 914 first determines whether there is second-stage second, third, and fourth color data based on the second-stage pattern flag acquired from the second-stage pattern flag storage unit 1403 (S1324). . For example, for the block of the second stage pattern flag “0”, it is determined that the second stage second, third, and fourth color data of the block of interest does not exist (No in S1324). For this reason, the data development unit 910 does not perform the pixel filling process using the second-stage second, third, and fourth color data for the target block, and proceeds to the next block process. If any other second-stage pattern flag is read, it is determined that the block has at least the second-stage second color data (Yes in S1324).

  If the data development unit 910 determines that the second stage second, third, and fourth color data exists for the target block (Yes in S1324), the data development unit 910 displays the unprocessed second stage in the target block. It is determined whether there are second, third, and fourth color data (S1325). If it is determined that there is unprocessed second-stage second, third, and fourth color data in the target block (Yes in S1325), the data expansion unit 910 displays the unprocessed second-stage second, third, and fourth color data. One of them is set as a processing target. Further, the data development unit 910 specifies the pixel position to be filled with the color data targeted for processing in the image data of 4 × 4 pixel size after development based on the second stage pattern flag and the first stage pattern flag. (S1326). Thereby, a 2nd specific means is implement | achieved.

  Since the data DMAC 913 sequentially transfers the second stage second, third, and fourth color data in one block to the data output unit 914, the data output unit 914 includes pixels to be filled with the second stage second color data, the third color The pixel to be filled with data and the pixel to be filled with the fourth color data are specified separately. Therefore, in S1326, it is assumed that the unprocessed color data is processed in order from the second color data in the second stage.

  The processing of S1324 to S1326 will be specifically described based on FIG. 14 and FIG. First, “2” is read as the second stage pattern flag, and it is determined that the arrangement pattern includes the second color data, as can be seen from FIG. That is, it is determined that the arrangement pattern when the second stage compression of the block of interest is performed is a pattern like the image data 1401. Accordingly, since only the second color data is stored in the second stage second, third, and fourth color storage unit 1405, the data output unit 914 uses the second color data received from the data DMAC 913 to determine which pixel. Calculate whether to fill. That is, the data output unit 914 determines that the pixel 02 and the pixel 20 are pixels to be filled with the second color data read from the data DMAC 913 based on the pattern “2” of the second stage pattern flag.

  Further, the pixel 02 and the pixel 20 are also the first color data of the upper right block and the lower left block of the image data 1101, respectively. Therefore, based on the first-stage pattern flag patterns “5” and “1” corresponding thereto, the data output unit 914 determines that the pixels 03, 13, 30, and 31 of the image data 1101 are also pixels to be filled. To do. That is, in the example of the compressed data in FIG. 14C, the second pixel of the pixels 02, 03, 13, 20, 30, and 31 shown above is converted into the second stage 2 in the flow of S 1324 to S 1326. The pixel position to be filled with color data is specified (FIG. 15B).

  Returning to the description of the processing flow in FIG. When the pixel position to be filled is specified, a data output unit is obtained from the start address of the development area storing the image data after raster development set in the data output unit 914 in advance, the size information of the developed image, and the position of the pixel to be painted. In step S1327, a memory address corresponding to a pixel to be filled with the second, second, third, and fourth color data is calculated. This implements a second memory address calculation unit.

  The data output unit 914 writes the color data to be processed in S1326 among the second, third, and fourth color data received from the data DMAC 913 in the memory area of the calculated address (S1328). This implements a second storage means. The processing of S1325 to S1328 is repeated for the number of pixels in which the second, third, and fourth color data are written in the target block. That is, in the example of the compressed data in FIG. 14C, since the second stage second color data is painted for 6 pixels as shown in FIG. 15B, the processing from S1325 to S1328 is performed six times.

  When the processing for one target block is completed in this way, in S1325, the data output unit 914 determines that there are no more pixels to be filled (No in S1325), and ends the processing for the target block. With the above processing, in the case of the compressed data in FIG. 14C, in the image data expanded on the memory, the pixel to which the second stage second color data is written is represented by the hatched portion in FIG. 15B. . In other words, the pixel value of the pixel represented by the hatched portion in FIG. 15B is expanded by the data expansion unit 910.

  Finally, the operation of the data expansion unit 920 that expands the first-stage second, third, and fourth color data will be described with reference to FIG. In the following description, the processing of S1341 to S1342 is described as being performed in parallel, but the processing can also be performed in order as shown in FIG. Note that when developing the first stage second, third, and fourth color data, the second stage pattern flag is not used, and therefore the pattern flag DMAC 922 may be omitted.

  The pattern flag DMAC 921 is set with the head address and image size of the first stage pattern flag storage unit 1402, reads the first stage pattern flag from the set address (0x00), and sequentially transfers it to the data output unit 914 ( S1341). In parallel, the data DMAC 923 is set with the head address and image size of the first stage second, third, and fourth color storage unit 1406, and the first stage second, third, and fourth colors from the set address (0x40). Data is read and sequentially transferred to the data output unit 924 (S1342).

  Based on the first stage pattern flag acquired from the first stage pattern flag storage unit 1402, the data output unit 924 first determines whether there is first stage second, third, and fourth color data (S1343). . For example, for the block of the first stage pattern flag “0”, it is determined that the first stage second, third, fourth color data of the target block does not exist (No in S1343). Therefore, the data development unit 920 does not perform the pixel filling process using the first-stage second, third, and fourth color data for the target block, and proceeds to the next block process. If any other first stage pattern flag is read, the data development unit 920 determines that the block has at least the first stage second color data (Yes in S1343).

  When the data development unit 920 determines that the first-stage second, third, and fourth color data exists for the target block (Yes in S1343), the data output unit 924 displays the unprocessed first in the target block. It is determined whether the second, third, and fourth color data are present (S1344). If it is determined that unprocessed first stage second, third, and fourth color data exists in the block of interest (Yes in S1344), the data output unit 924 displays the unprocessed first stage second, third, and fourth color data. One of the data is processed. Further, the data output unit 924 specifies a pixel position to be filled with the color data to be processed in the developed image data based on the first stage pattern flag (S1345).

  Since the data DMAC 923 sequentially transfers the first, second, third, and fourth color data in one block to the data output unit 924, the data output unit 924 includes the pixels to be filled with the first step second color data, the first color The pixels to be filled with the first stage third color data and the pixels to be filled with the first stage fourth color data are specified separately.

  The processing of S1343 to S1345 will be specifically described based on FIG. 14 and FIG. First, when the pattern “B” is first read as the first stage pattern flag, it is determined that the arrangement pattern includes the second color data and the third color data, as can be understood with reference to FIG. The That is, the arrangement pattern when the first stage compression of the block of interest is first determined to be a pattern like the upper left block of the image data 1101. Therefore, since the second color data and the third color data are stored in the first stage second, third, and fourth color storage unit 1406, the second color data and the third color data received from the data DMAC 923 are Which pixels are filled is calculated. That is, the data output unit 924 determines that the pixel 01 and the pixel 10 are pixels to be filled with the second color data read from the data DMAC 923 based on the pattern “B” of the first stage pattern flag. Similarly, the pixel 11 is determined to be a pixel to be filled with the third color data read from the data DMAC 923.

  Returning to the description of the flow in FIG. When the pixel position to be filled is specified, the data output unit 924 determines the first address from which the rasterized image set in advance in the data output unit 924 is stored, the size information of the developed image, and the position of the pixel to be filled. A memory address corresponding to a pixel to be painted with the first, second, third, and fourth color data is calculated (S1346).

  Then, the data output unit 924 transfers the color data to be processed in S1345 out of the second, third, and fourth color data received from the data DMAC 923 to the memory area of the calculated address and writes it (S1347). The processing of S1344 to S1347 is repeated for the number of pixels in which the first stage second, third, and fourth color data are written in the target block.

  When the processing for one block is completed in this way, in S1344, the data output unit 924 determines that there are no pixels to be filled (No in S1344), and ends the development processing for the block of interest. With the above processing, in the case of the compressed data in FIG. 14C, in the image data expanded on the memory 106, the pixels in which the first, second, third, and fourth color data are written are in each expanded block. It is represented by the hatched portion in FIG. In other words, the pixel value of the pixel represented by the hatched portion in FIG.

[Contrast in number of accesses]
The development unit 211 of the present invention is characterized in that it can be arranged in parallel like the data development units 900, 910, and 920, and can be operated independently, and the memory access can be made efficient.

  Regarding the memory access amount necessary for the decompression processing of the two-stage compressed data, when the decompressing unit 211 is configured based on the above-described present invention (FIG. 9), the configuration as illustrated in FIG. 16 (a) and 16 (b) respectively show the case where it is configured to expand in stages in the reverse order of compression (hereinafter referred to as the conventional configuration). FIG. 20 shows a processing flow as a conventional configuration. FIG. 16A shows the memory access amount when the data (FIG. 14B) obtained by compressing the image data 1101 in two stages is expanded into raster data by the expansion unit (FIG. 9) of the present invention. FIG. 16B shows the memory access amount when the data (FIG. 14B) obtained by compressing the image data 1101 in two stages is developed into raster data by the development unit (FIG. 19) having the conventional configuration. . As a difference in access type between the present invention shown in FIG. 16A and the conventional configuration shown in FIG. 16B, in the present invention, “first stage first color data read (corresponding to 1613 in FIG. 16B)” is shown. ) ”Is not necessary, and the item is missing.

  First, the expansion unit 211 of the present invention converts the second stage first color data, the second stage second, third, fourth color data, and the first stage second, third, fourth color data into data DMACs 903, 913, and 923, respectively. Read. The total amount of data read is 3 bytes (1603), 3 bytes (1604), and 21 bytes (1602) in the case of the compressed data example in FIG. The first stage pattern flag is read by pattern flags DMAC 901, 911, and 921, respectively, and the second stage pattern flag is read by pattern flags DMAC 902 and 912. In the example of compressed data shown in FIG. 14C, the amount of data to be read is 2 bytes for 3 modules (1600) and 1 byte for 2 modules (1601), respectively. In addition, since the output data is written in a total of the data output units 904, 914, and 924 for the amount of raster data of the image data 1101, it corresponds to 16 pixels (1605) of 3 bytes. The total memory access amount (read and write data amount) is 83 bytes.

  On the other hand, in the case of the expansion unit 1900 having the conventional configuration, firstly, the first stage of expanding the second-stage compressed data into the first-stage compressed data, and then the next stage of expanding the first-stage compressed data into raster data. It will be divided into two. Therefore, in the first stage of development, the second stage pattern flag, the second stage first color data, and the second stage second, third, and fourth color data are read. In the case of the compressed data example in FIG. 14C, the respective memory access amounts are 1 byte (1611), 3 bytes (1614), and 3 bytes (1615). Further, the first-stage compressed data after the second-stage compressed data is expanded by one stage is as shown in FIG. Therefore, in the next development stage where the first stage compressed data is developed into raster data, the first stage pattern flag, the first stage first color data, and the first stage second, third, and fourth color data are read. . In the case of the compressed data example of FIG. 11C, the respective memory access amounts are 2 bytes (1610), 3 bytes of 4 pixels (1612), and 21 bytes (1613).

  The output data is written in two stages, a process of writing the result of expanding the second-stage compressed data into the first color storage unit 1103 and a process of writing raster data as a result of expanding the first-stage compressed data. These are equivalent to data output of raster data amounts of the image data 1401 and the image data 1101, respectively. Therefore, each memory access amount is 4 pixels of 3 bytes and 16 pixels of 3 bytes (1616). Therefore, the total memory access amount (read and write data amount) is 102 bytes. In the conventional configuration, the division of roles for each part that accesses the memory as in the present invention is not specified, and it is assumed that the development unit 1900 generally performs the division of roles.

  The development unit in FIG. 9 reads and uses the same pattern flag by a plurality of data development units. Accordingly, the memory access amount related to the pattern flag is increased more than that of the expansion unit in FIG. 19 (1600, 1601, 1610, 1611). On the other hand, in this embodiment, the two-stage compressed data is directly expanded into raster data and written back to the memory. Therefore, part of the data writing back to the memory and the reading of the compressed data to the first stage when the compressed data of the second stage are expanded to the compressed data to the first stage and output (1612, 1605, 1616). Since the data amount of the pixel data is larger than the data amount of the pattern flag, the total memory access amount according to the present embodiment in FIG. 9 is larger than the memory access amount in the case of developing in stages as shown in FIG. It can be seen that it is reduced by 20%.

  In addition, a plurality of expansion units shown in FIG. 19 are prepared for the expansion of the second-stage compressed data into the first-stage compressed data and the expansion of the first-stage compressed data into the raster data, and an intermediate buffer is provided therebetween. With this configuration, the number of memory accesses to the system memory can be reduced. However, such a configuration increases the circuit scale. On the other hand, since the data expansion units 900, 910, and 920 according to the present embodiment in FIG. 9 can directly expand the two-stage compressed data into a raster image, such an intermediate buffer is not necessary.

  Further, when handling compressed data of three or more stages, if a decompression unit 1900 shown in FIG. 19 is used, a decompression unit is required at each stage of data decompression. Requires an expansion part and an intermediate buffer. For example, when decompressing data compressed in three stages, as shown in FIG. 17A, many decompressing units (1 + 4 + 16) and intermediate buffers (4 + 16) between them are required. On the other hand, in the present invention, one data expansion unit may be added to the data expansion units 900, 910, and 920 of FIG. 9 as shown in FIG.

  More specifically, as shown in FIG. 18, each data compression unit is changed so as to be able to read the pattern flag in the three stages in order to expand the three-stage compressed data. The internal algorithm in the data output unit may increase the determination of the fill pixel position by the pattern flag by one step. By making such a change, the third stage first color data, the third stage second, third, fourth color data, the second stage second, third, fourth color data, the first stage second, third, fourth color data Are processed by modules (data development units) 1801, 1802, 1803, and 1804, respectively. As a result, the expansion to the expansion of the third stage compressed data becomes possible while suppressing the addition of the intermediate buffer.

  That is, when decompressing compressed image data that has been subjected to N stages of compression processing, it is sufficient to provide N + 1 data decompressing units. With this multiplexed configuration, the representative color decoding means for expanding the representative color obtained by the Nth compression process using the pattern flag obtained by the 1st to Nth encoding, and the 1st to Mth ( N is an interpolation color decoding means for developing using the interpolation colors obtained by each of the compression processes (M is an integer from 1 to N).

  In the development unit of the present invention, depending on the image, processing may be concentrated on some data development units. In such a situation, if each data development unit has the same configuration, it can function as a data development unit that can be used in any data development unit, and may be configured to distribute the load. Absent.

[Second Embodiment]
In the first embodiment, when outputting data obtained by compressing each block, first color data is always output in addition to the pattern flag, and is written to the memory. In the compression mode described in the present invention, in addition to the definition of the pattern flag of FIG. 4, it is also possible to define and use a pattern flag “F” called a repeat flag. This is because the repeat flag “F” is output as a pattern flag for a block having the same color arrangement (arrangement pattern) as the block processed once before the block compression. The second, third, and fourth color data are not output to the memory.

  As a result, the amount of data output to the memory after compression can be further reduced. Note that the previous block here is a processed block adjacent to the block of interest when, for example, block processing is performed in raster order. At this time, since the processing order is a reference, the position may be any position with respect to the block of interest, for example, located on the block of interest. In addition, when a repeat flag is added to a pattern flag, the pattern flag is represented by 4 bits in the present embodiment. However, even if one repeat flag is added in addition to 15 normal pattern flags, the data amount of the pattern flag There is no change in (4 bits).

  In the second embodiment, a decompression flow of compressed data in which repeat flags are mixed will be described. Further, in the second embodiment, at the time of compression, a repeat flag is added only in the final compression stage, and a repeat flag is not added to the head block of the main scanning line. That is, when the image data 2100 shown in FIG. 21 is taken as an example in the case where the image data 2100 shown in FIG. On the other hand, if this is the second stage compression (final stage compression) and there is a block having the same color arrangement as the left adjacent block, the repeat flag “F” is set in the pattern flag for that block. For example, in the second stage compressed block 2102 of FIG. 21, it is assumed that the block described as “R” is a block in which the repeat flag “F” is set. That is, the repeat flag “F” means that the block S2_04 is a repetition of the block S2_00, and the block S2_44 is a repetition of the block S2_40. Here, it is determined whether the block S2_44 of the second stage compressed block is a block to which a repeat flag should be added based on the first color generated during the first stage compression. That is, the block S2_40 and the block S2_44 are blocks in the image data 2101 composed of the first color after the first stage compression, although the original image pattern is different, as shown in the image data 2100 of FIG. Since the same color arrangement is set, a repeat flag is set in the block S2_44. The state of the compressed data on the memory when the second-stage compressed block is formed is expressed as shown in FIGS.

  In the image data 2100 to be processed in the first stage compression, if the pixel value of the upper left pixel of each block is the first color data, the first stage second, third, and fourth color data are the pixels 23, 27, 43, Pixel values of 45, 47, 63, 65, and 67 are stored in the first stage second, third, and fourth color storage unit 2205. In addition, the pattern flag at the time of the first stage compression is stored in the first stage pattern flag storage unit 2201. In the image data 2101 to be processed in the second stage compression, the pixel value of the upper left pixel of each block is the first color, but the upper left block and the upper right block have the same color arrangement, and the lower left block and The lower right block has the same color arrangement. Accordingly, in the second stage first color data, only the pixel values of the pixels 00 and 40 are stored in the second stage first color storage unit 2203, and the pixels 04 and 44 of the block having the same color and the repeat flag are given. No pixel value is stored.

  In the second stage second, third, and fourth color data, the pixel value of the pixel 22 is stored in the second stage second, third, and fourth color storage unit 2204, but the pixel of the pixel 26 of the block to which the repeat flag is added. The value is not stored in memory. Further, the pattern flag at the time of the second stage compression is stored in the second stage pattern flag storage unit 2202, and a repeat flag “F” is given to the blocks determined to have the same color arrangement at this time.

  The operation when the compressed data compressed in multiple stages in this way is expanded by the expansion unit 211 of FIG. 9 will be described. The flow will be described in detail only for the differences based on FIG. 13 that has already been described. Here, each data development unit 900, 910, and 920 has a buffer for holding the pixel value of the previously processed block for the repeat flag processing.

  First, the data development unit 900 that develops the second stage first color determines whether or not the second stage pattern flag is a repeat flag at the time of specifying the fill pixel in S1304. The normal processing described in the embodiment is performed. If it is a repeat flag, the data expansion unit 900 operates as follows.

  When receiving the repeat flag, the data development unit 900 that performs the second stage first color development uses the first color data of the previous block stored in the buffer as the transfer data in S1307, not the data received from the data DMAC 903. To do. When receiving the repeat flag, the data expansion unit 910 that expands the second, second, third, and fourth colors, instead of the data received from the data DMAC 913, receives the second, third, and fourth colors of the previous block stored in the buffer. The data is used as transfer data in S1328. On the other hand, the data development unit 920 that develops the first-stage second, third, and fourth colors does not read the second-stage repeat flag and does not perform processing dependent on the second-stage repeat flag, and thus performs the same processing as in the first embodiment.

  Through the above processing, the decompression unit 211 of FIG. 9 can perform decompression processing of compressed data even for compressed data in which repeat flags are mixed. Thereby, the effect similar to 1st embodiment can be acquired.

<Other embodiments>
The present invention can also be realized by executing the following processing. That is, software (program) that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media, and a computer (or CPU, MPU, or the like) of the system or apparatus reads the program. It is a process to be executed.

Claims (5)

The image data is divided into blocks each having a 2 × 2 pixel size, the pixel value of a pixel at a predetermined position in each of the divided blocks is a representative color, the pixel values other than the representative color in each block are interpolated colors, The arrangement pattern of the representative color and the interpolation color in the block is acquired as arrangement information, and the representative means, the interpolation color, and the arrangement information are stored in different storage areas, respectively, by encoding means N−1th (N Is an image processing apparatus that decodes compressed data obtained by performing the N-th encoding on image data composed of representative colors obtained by encoding (an integer of 2 or more),
The position where the representative color obtained by the Nth encoding by the encoding means is to be developed is specified based on the arrangement information obtained by the 1st to Nth encodings by the encoding means, and the identification Representative color decoding means for expanding the representative color at the position,
The position where the interpolation color obtained by the Mth encoding (M is an integer from 1 to N) by the encoding unit is to be developed is obtained by the 1st to Mth encoding by the encoding unit. N interpolating color decoding means for identifying each of the arranged arrangement information and expanding the interpolated color obtained by the M-th encoding at the identified position,
The image processing apparatus, wherein the representative color decoding unit and the N interpolation color decoding units operate independently of each other. The representative color decoding means includes:
First specifying means for specifying the position where the representative color in the developed image is to be developed using each of the arrangement information acquired by encoding N times by the encoding means;
First memory address calculation means for calculating a memory address of a development area in which data of the image after development corresponding to a position where the representative color specified by the first specification means is to be developed;
The image processing apparatus according to claim 1, further comprising: a first storage unit that stores the representative color at a memory address of a development area calculated by the first memory address calculation unit. Interpolation color decoding means for expanding the interpolation color obtained by the M-th encoding,
Using each of the arrangement information acquired by encoding 1 to M times by the encoding means, a second specifying means for specifying the position where the interpolation color in the developed image is to be developed;
Second memory address calculation means for calculating a memory address of a development area in which data of the image after development corresponding to a position where the interpolation color specified by the second specification means is to be developed;
3. The second storage means for storing the interpolation color obtained by the M-th encoding at the memory address of the development area calculated by the second memory address calculation means. The image processing apparatus described. The image data is divided into blocks each having a 2 × 2 pixel size, the pixel value of a pixel at a predetermined position in each of the divided blocks is a representative color, the pixel values other than the representative color in each block are interpolated colors, The arrangement pattern of the representative color and the interpolation color in the block is acquired as arrangement information, and the representative means, the interpolation color, and the arrangement information are stored in different storage areas, respectively, by encoding means N−1th (N Is an image processing method for decoding the compressed data obtained by performing the Nth encoding on the image data composed of the representative color obtained by the encoding of 2 or an integer),
The representative color decoding unit determines the position where the representative color obtained by the Nth encoding by the encoding unit should be developed based on the arrangement information obtained by the first to Nth encodings by the encoding unit. A representative color decoding step of developing the representative color at the specified position;
Each of the N interpolation color decoding means uses the encoding means to determine the position where the interpolation color obtained by the M-th encoding (M is an integer from 1 to N) by the encoding means should be developed. An interpolation color decoding step that specifies each of the arrangement information obtained by the 1st to Mth encodings and develops the interpolation color obtained by the Mth encoding at the specified position,
The image processing method, wherein the representative color decoding step executed by the representative color decoding unit and the interpolation color decoding step executed by the N interpolation color decoding units are performed independently. The image data is divided into blocks each having a 2 × 2 pixel size, the pixel value of a pixel at a predetermined position in each of the divided blocks is a representative color, the pixel values other than the representative color in each block are interpolated colors, The arrangement pattern of the representative color and the interpolation color in the block is acquired as arrangement information, and the representative means, the interpolation color, and the arrangement information are stored in different storage areas, respectively, by encoding means N−1th (N In order to decode the compressed data obtained by performing the N-th encoding on the image data composed of the representative colors obtained by the encoding of
Computer
The position where the representative color obtained by the Nth encoding by the encoding means is to be developed is specified based on the arrangement information obtained by the 1st to Nth encodings by the encoding means, and the identification Representative color decoding means for expanding the representative color at the position,
The position where the interpolation color obtained by the Mth encoding (M is an integer from 1 to N) by the encoding unit is to be developed is obtained by the 1st to Mth encoding by the encoding unit. And based on each of the arrangement information, function as N interpolation color decoding means for expanding the interpolation color obtained by the M-th encoding at the specified position,
The representative color decoding unit and the N interpolation color decoding units operate independently of each other.

Images