JP6675253B2 - Image decoding apparatus and method, and image processing apparatus - Google Patents

Description

  The present invention relates to a process for decoding compressed image data.

  An image compression technique is applied to high-resolution image data having an enormous data amount in order to store the image data with less memory resources or to transfer the image data to a network in a shorter time. On the other hand, image compression technology is required not to impair image quality. Patent Literature 1 discloses an image compression technique for reducing an amount of data and reducing image quality by dividing an image into one or more regions each including one or more pixels. A technique for switching is disclosed.

  As a technique for dividing an image into one or more regions and switching a data compression method, image data is encoded into region boundary data (hereinafter, region boundary data) and region compressed data (hereinafter, region compressed data). There are technologies to do this. The region boundary data includes information for specifying the coordinates of the boundary, and data compression method and attribute information for specifying the region compressed data for the region on the right side of the boundary.

  The above technique employs lossless compression run-length encoding and lossy compression JPEG as data compression schemes. Therefore, there is a restriction that the area to which the lossy compression method is applied (hereinafter, lossy compression area) must be an integral multiple of the minimum coding unit (MCU) composed of eight pixels both horizontally and vertically. For this reason, the lossy compression area includes unnecessary pixels overlapping with the area to which the lossless compression method is applied (hereinafter, lossless compression area).

  Therefore, an apparatus that decodes code data in which a lossy compression area and a lossless compression area coexist performs the following two steps. In a first step, data in a lossy compression area (hereinafter, lossy compression data) is decoded and written to a memory. In the next second step, the coordinates of the boundary are specified, and the data of the lossless compressed area (hereinafter, lossless compressed data) is decoded and written to the memory. At this time, in the first step, the data is written to the lossless compressed area. Overwrite the pixel data.

  As described above, when writing the decompression result of the compressed image having the lossy compression data and the lossless compression data to the memory, there are pixels that overwrite the decompression result of the lossy compression data with the decompression result of the lossless compression data. As a result, the system bus band and the memory band are consumed extra.

U.S. Patent No. 5,883,976

  The present invention has been made in view of the above problems. Then, when the first image data compressed by the first method and the second image data compressed by the second method are mixed, the present invention provides the second image in the decompression result of the first image data. A technique for decoding image data without overwriting the decompression result of the image data.

  The present invention has the following configuration as one means for achieving the above object.

First data obtained by compressing an image by a first method for each area and second data obtained by compressing a partial area of the image by a second method different from the first method are mixed. An image decoding device that decodes data and outputs the image,
First decoding means for decoding the first data and outputting pixel data of each pixel constituting an image;
Second decoding means for decoding the second data in the compressed area based on boundary information indicating a compressed area compressed by the second method;
Writing means for writing pixel data of each pixel constituting the image obtained from the second decoding means to a memory,
The first decoding means outputs pixel data of each pixel to the second decoding means,
The second decryption means includes:
Determine whether the processing target pixel is the compression area based on the boundary information,
Pixel data obtained from the first decoding unit is output to the writing unit for pixels other than the compressed region, and pixel data obtained by decoding by the second decoding unit for pixels included in the compressed region. Is output to the writing means.

  According to the present invention, when the first image data compressed by the first method and the second image data compressed by the second method are mixed, the second image data is decompressed in the first image data. The image data can be decoded without overwriting the decompression result of the image data.

The figure which shows the memory image of source bitmap data. The figure which shows an example of the rectangular area of 16x16 pixels. FIG. 6 is a diagram showing a format example of boundary information of a rectangular area and lossless compressed data. FIG. 4 is a diagram for explaining lossless compressed data of a rectangular area. FIG. 2 is a block diagram illustrating a configuration example of an image decoding device according to the embodiment. FIG. 3 is a block diagram illustrating a configuration example of a lossless decoding unit. 9 is a flowchart for describing pixel data output processing, including decoding processing of a lossless decoding unit. 9 is a flowchart illustrating a process of acquiring a subsequent boundary. 9 is a flowchart for describing processing for updating a processing area and a continuation boundary queue when a line is not completed. 9 is a flowchart for describing processing for updating a processing area and a continuation boundary queue when a line ends. 9 is a flowchart for explaining a lossless compression area decoding process. The figure which shows an example of bitmap data. FIG. 2 is a block diagram illustrating a system configuration example of an image processing device including an image decoding device. FIG. 4 is a diagram for explaining each parameter of a lossless decoding unit. FIG. 13 is a block diagram showing a modification of the image decoding device. FIG. 2 is a block diagram illustrating a configuration example of an image processing device having a rendering system to which the image decoding device is applied.

  Hereinafter, an image decoding apparatus and an image decoding method according to an embodiment of the present invention will be described in detail with reference to the drawings. It should be noted that the embodiments do not limit the present invention according to the claims, and all combinations of the configurations described in the embodiments are not necessarily essential to the solution of the present invention. In the following, the description will be given assuming that the scanning direction of the pixels is a raster scan from left to right and from top to bottom, but there is no limitation on the scanning direction.

[Structure of code data]
Before describing the image decoding apparatus, the structure of code data will be described. FIG. 1 shows a memory image 10 of source bitmap data. The memory image 10 is bitmap data included in a drawing area such as a page or a frame, and includes a bitmap header 11 and compressed data 12.

  The bitmap header 11 is a set of headers defined for each bitmap (bitmaps 1, 2,..., B) included in a drawing area such as a page or a frame. For example, the header of the bitmap 1 includes a bitmap size 111, a channel number 112, a bit precision 113, a pointer map 114 which is a set of pointers of lossless compressed data, and a pointer map 115 which is a set of pointers of lossy compressed data. It is. Hereinafter, the lossless compressed data may be referred to as “lossless data”, and the lossy compressed data may be referred to as “lossy data”.

  The compressed data 12 is a set of compressed data for each bitmap (bitmaps 1, 2,..., B) included in a drawing area such as a page or a frame. , WH lossless data 117, 118,..., 119, and rectangular areas 1, 2,..., WH lossy data 120, 121,. Includes 122.

  Each bitmap is divided into rectangular areas of m × m pixels (for example, 64 × 64 pixels). For example, bitmap 1 includes W × H rectangular areas. If the size of the bitmap is not an integer multiple of m × m pixels, padding is added outside the bitmap so as to be an integer multiple of m × m pixels, and then the bitmap is divided into rectangular regions, Data compression is performed on the rectangular area.

  The bitmap size 111 indicates how many times the width and height of the corresponding bitmap are larger than the rectangular area of m × m pixels. The channel number 112 indicates the number of color components. RGB data has three channels, CMYK data has four channels, YUV data has three channels, and monochrome (BW) has one channel. The bit precision 113 indicates the bit depth per channel.

  The pointer map 114 is a set of pointers of the lossless data, and each pointer indicates the head address of the lossless data in the corresponding rectangular area. The pointer map 115 is a set of pointers to lossy data, and each pointer indicates the start address of the lossy data in the corresponding rectangular area.

  The lossless data 117, 118,..., 119 are data obtained by losslessly compressing at least a part of the image data of each rectangular area of the bitmap 1. Each lossless data includes boundary information in a corresponding rectangular area and color information of a single color area.

  The lossy data 120, 121,..., 122 are data obtained by lossy compression of the image data of each rectangular area of the bitmap 1. Each lossy data is a set of compressed data for each minimum coding unit (MCU) of the corresponding rectangular area. For example, JPEG is used for the lossy compression of the embodiment. Assuming that the MCU of JPEG is 8 × 8 pixels and the rectangular area is 64 × 64 pixels, each lossy data is a set of JPEG data corresponding to each of 8 × 8 = 64 MCUs.

FIG. 2 shows an example of a rectangular area 20 of 16 × 16 pixels when m = 16. In FIG. 2, a thick solid line indicates an end (boundary of a pixel set) of the rectangular area 20, and horizontal and vertical broken lines indicate a boundary of 8 × 8 pixels corresponding to a JPEG MCU. The rectangular area 20 includes, for example, the following areas.
First area 21: lossless compression area,
Second area 22: lossy compression area,
Third area 23: lossless compression area,
Fourth area 24: lossless compression area,
Fifth area 25: lossy compression area,
Sixth area 26: lossless compression area,
Seventh area 27: lossless compression area,
Eighth area 28: Lossy compression area.

  In the rectangular area 20, adjacent pixels of the same color are divided into lossless compression areas. However, JPEG used in the embodiment is a method of encoding for each predetermined rectangular area. Therefore, in the case of the rectangular area 20 shown in FIG. 2, compression is performed by two compression methods as follows. Of the data in the rectangular area 20, data of pixels in the lossless compression area are encoded by a lossless compression method (hereinafter, lossless encoding). After replacing all the pixels in the lossless compression area with predetermined pixel values set in consideration of the compression efficiency, the data in the rectangular area 20 is encoded by a lossy compression method (hereinafter, lossy encoding).

  A format example of the boundary information of the rectangular area 20 and the lossless compressed data 41 will be described with reference to FIG. FIG. 3 (A) shows the boundary information. The bold solid line corresponds to the attribute “start boundary” indicating the start pixel of the lossless compression area on the horizontal line (hereinafter, “line”), and the bold broken line is the lossless compression area on the line. Corresponding to the attribute “termination boundary” indicating the termination pixel of. That is, a certain lossless compression area starts with the pixel at the start boundary and ends with the pixel at the end boundary for each line.

The start boundary and the end boundary are generated by the encoding device so as not to cross each other. The start boundary and the end boundary are generated as a pair of boundary information. In the lossless compression area, the correspondence between each pair of the start boundary and the end boundary shown in FIG. 3A and the area shown in FIG. 2 is as follows.
Start boundary 31 and end boundary 32: first region 21,
Start boundary 33 and end boundary 34: third region 23,
Start boundary 35 and end boundary 36: fourth region 24,
Start boundary 37 and end boundary 38: sixth region 26,
Start boundary 39 and end boundary 40: seventh region 27.

  FIG. 3B shows a format example of the lossless compressed data 41. The lossless compressed data 41 corresponds to the lossless compressed data of the rectangular area 20. The boundary number 411 indicates the number E of pairs of the start boundary and the end boundary in the rectangular area 20. The header 412 at the boundary 0 is header information of the start boundary and the end boundary of the first lossless compression area. The header 413 of the boundary E-1 is header information of the start boundary and the end boundary of the E-1st lossless compression area.

The trajectory data 414 of the boundary 0 is data for specifying the horizontal coordinate X (hereinafter, coordinate X) of the boundary 0, and may use a function, a coordinate X of each line, a difference between the coordinates X of each line, and the like. it can. The locus data 415 of the boundary E-1 is data for specifying the coordinates X of the boundary E-1. The edge header 42 indicates the format of each of the header information 412 and 413, and stores the following values.
Region color (RegionColour) 421: Color of lossless compression region,
Start Y-Position 422: Vertical coordinate Y (hereinafter, coordinate Y) of the start point of the start boundary,
Starting X-Position 423: Horizontal coordinate X of the start point of the start boundary,
End Y-Position 424: The vertical coordinate Y of the end point of the start boundary,
Disabling X-Position 425: Horizontal coordinate X of the start point of the end boundary.

  Note that the coordinates Y of the start point and the end point of the end boundary are equal to the coordinates Y of the start point and the end point of the paired start boundary. The boundary headers 412 and 413 are sorted in ascending order of the leading Y coordinate 422. Headers having the same start Y coordinate 422 are sorted in ascending order of the start X coordinate 423. In other words, the boundary headers are arranged in the order in which the rectangular area 20 is scanned from upper left to lower right, and stored in the memory.

  In each line, from the coordinate X of the start boundary indicating the start information of the lossless compression area to the coordinate X of the end boundary indicating the end information of the lossless compression area, pixels of the color defined as the area color 421 are obtained. If there is no gap between the end boundary and the next start boundary on one line, it indicates that the lossless compression regions are adjacent. Also, if there is a gap between the end boundary and the next start boundary on one line, there is a lossy compression region between the lossless compression regions, and the area from the right of the end boundary to the left of the next start boundary is , Pixels obtained by expanding the lossy compressed data.

  If there is no start coordinate at the left end of the line (X = 0), the lossy compressed data is expanded from the left end of the line (X = 0) to the left of the first appearing start coordinate X. The resulting pixel. If there is no start coordinate on the line, the line is a lossy compression area, and the pixels from the left end (X = 0) to the right end (for example, X = 15) of the line are obtained by expanding the lossy compressed data. become.

  When the data compression method of the image data in the lossy compression area of the rectangular area 20 is JPEG, as described above, the lossy compression data is composed of a total of four MCUs in which two 8 × 8 pixel MCUs are vertically and horizontally arranged two by two, Includes all pixels in the rectangular area 20. Therefore, when the lossy compressed data is expanded, pixels overlapping with the lossless compression areas 21, 23, 24, 26, and 27 shown in FIG. 2 are decoded. However, it should be noted that the pixel values of the lossless compression areas 21, 23, 24, 26, and 27 obtained by decoding the lossy compressed data are predetermined replaced pixel values.

  The lossless compressed data of the rectangular area 20 will be described with reference to FIG. In the following description, it is assumed that the number of channels of the color component is 3, and the color space is RGB. Since the rectangular area 20 has five pairs of start and end boundaries corresponding to the lossless compression areas 21, 23, 24, 26, and 27, the number of boundaries E = 5. The boundary 0 corresponds to the first area 21. In the header at the boundary 0, as shown in FIG. 4, the area color is set as “0xFF0000”. That is, the first area 21 is composed of pixels of colors R = 255, G = 00, and B = 00. Also, the start Y coordinate “0” of the start boundary 31 and the end boundary 32, the start X coordinate “0” of the start boundary 31, the end Y coordinate “9” of the start boundary 31 and the end boundary 32, and the end X coordinate of the end boundary 32 "2" is set.

  Similarly, boundary 1 corresponds to third region 23, boundary 2 corresponds to fourth region 24, boundary 3 corresponds to sixth region 26, boundary 4 corresponds to seventh region 27, and boundary 1 corresponds to boundary 7. Data shown in FIG. 4 is set in each header of No. 4.

  The locus of the start boundary of boundary 0 is between line Y = 0 and line Y = 9, and the X coordinate is 0, 0, 0, 0, 0, 0, 1, 2, in the ascending order of the Y coordinate of the line. 3 and 4. As the trajectory data of the start boundary of the boundary 0, a difference “0, 0, 0, 0, 0, +1, +1, +1, +1” of the X coordinate between the lines in the scanning order is set. The trajectory of the end boundary of boundary 0 is between line Y = 0 and line Y = 9, and the X coordinate is 2, 5, 5, 6, 7, 6, 6, 6, 5, and 5. As the trajectory data of the end boundary of the boundary 0, a difference “+3, 0, +1, +1, −1, 0, 0, −1, 0” of the X coordinate between the lines in the scanning order is set. In this way, a non-linear boundary of the first region 21 can be defined.

The boundary 1 corresponds to the third area 23. Therefore, the trajectory of the start boundary and the end boundary of boundary 1 is between line Y = 0 and line Y = 11.
The boundary 2 corresponds to the fourth area 24. Therefore, the trajectory of the start boundary and the end boundary of boundary 2 is between line Y = 2 and line Y = 9.
The boundary 3 corresponds to the sixth area 26. Therefore, the trajectory of the start boundary and the end boundary of boundary 3 is between line Y = 7 and line Y = 13.
The boundary 4 corresponds to the seventh region 27. Therefore, the trajectory of the start boundary and the end boundary of boundary 4 is between line Y = 11 and line Y = 13.
As described above, the trajectory data as shown in FIG. 4 is set.

  Hereinafter, by referring to the boundary information, a decoding process of decoding image data in which image data compressed by the lossy compression method and image data compressed by the lossless compression method are mixed without overwriting pixel data is described. explain.

[First embodiment]
[Configuration of Image Decoding Device]
FIG. 5 is a block diagram showing an example of the configuration of the image decoding apparatus according to the embodiment. The control unit 51 reads the source bitmap data stored in the memory 55 and the buffer information indicating the address and size of the buffer area for writing the decoded pixel. The control unit 51 interprets the bitmap header 11 and controls the lossy decoding unit 52, the lossless decoding unit 53, and the writing processing unit 54. Since the source of the source bitmap data to be decoded stored in the memory 44 does not matter, it is not shown.

  Under the control of the control unit 51, the lossy decoding unit 52 reads the lossy compressed data stored in the memory 55, decodes the lossy compressed data, and outputs the decoded pixel data to the lossless decoding unit 53. The lossy decoding unit 52 includes, for example, a JPEG decoder and a buffer 56 for 16 × 8 pixels, performs rearrangement processing of decoding results of the vertical 1 × horizontal 2 MCU, and converts pixel data in the rectangular area into 16-pixel lines. , And sequentially output to the lossless decoding unit 53. The lossy decoding unit 52 validates the state in which the pixel data is stored in the buffer 56. In this embodiment, the lossy decoding unit 52 is realized by a dedicated circuit. Each pixel data output from the lossy decoding unit 52 is output to the lossless decoding unit 53 without being stored in the memory 55.

  The lossless decoding unit 53 generates pixel data of a rectangular area by decoding the lossless compressed data stored in the memory 55 under the control of the control unit 51. The lossless decoding unit 53 receives the pixel data output from the lossy decoding unit 52, selects one of the pixel data output from the lossy decoding unit 52 and the pixel data obtained by decoding the lossless compressed data for each pixel, and writes it. Output to the output processing unit 54.

  In this embodiment, the lossless decoding unit 53 is provided separately from the CPU, and is realized by software processing in which a processor having a dedicated work memory reads and executes a program from a dedicated program memory or a shared system memory. And The processor that realizes the lossless decoding unit 53 has one or more interfaces for performing data transmission and reception with other components. In particular, an interface for connecting the lossless decoding unit 53 and the lossy decoding unit 52 includes, from the lossy decoding unit 52, a valid signal for notifying the above-described validity, and the lossless decoding unit 53 transmitting pixel data from the buffer 56 of the lossy decoding unit 52. It has a ready signal indicating a ready state to be read. Further, the interface has a data signal for performing data transfer from the lossy decoding unit 52.

  Under the control of the control unit 51, the writing processing unit 54 calculates an address to which the pixel data output from the lossless decoding unit 53 is to be written, and writes the pixel data to the buffer area of the memory 55. In the present embodiment, the write-out processing unit 54 is also realized by a processor that realizes the lossless decoding unit 53 by software processing that reads and executes a program.

[Lossless Decoding Unit] A configuration example of the lossless decoding unit 53 is shown in the block diagram of FIG. The decoding control unit 73 controls each component of the lossless decoding unit 53, which will be described later, and performs an image decoding process. The details of the image decoding process will be described later.

The boundary reading unit 61 includes a new boundary buffer for temporarily storing boundary data read from the memory 55 as “new boundary” (or new boundary information) described later, and a boundary as a counter indicating the number of unprocessed new boundaries. It has a remaining number R _N. Boundary reading unit 61 reads the number of boundaries from lossless compression data stored in the memory 55 E (see FIG. 4), sets the read boundary number E in the boundary remaining number R _N is a counter. Also, a boundary header and a set of corresponding trajectory data (hereinafter, boundary data) are read, and the read boundary data is stored in a new boundary buffer. When the boundary data from the new boundary buffer is outputted, the boundary remaining number R _N is decremented, the boundary data is deleted from the new boundary buffer. Hereinafter, a pair of the start boundary and the end boundary indicated by the boundary data stored in the new boundary buffer is referred to as “new boundary” (or new boundary information).

  The continuous boundary queue 62 includes a FIFO into which boundary data selected as “subsequent boundary” (or subsequent boundary information) described later is input, and a continuous boundary buffer that temporarily stores boundary data extracted from the FIFO. Hereinafter, a pair of the start boundary and the end boundary indicated by the boundary data stored in the continuation boundary buffer is referred to as “continuation boundary” (or continuation boundary information).

Further, the continuation boundary queue 62 stores, as variables for managing the number of boundary data in the FIFO, a boundary remaining number R _Lc which is a counter indicating the number of boundaries on the line of interest and the number of boundaries on the next line. It has a boundary remaining number C _Ln which is a counter shown. Although details will be described later, the boundary remaining number R _Lc is decremented when boundary data is read from the continuous boundary buffer. The boundary remaining number C _Ln is incremented when boundary data is input to the FIFO. Then, when the line of interest is incremented, the boundary remaining number C _Ln is copied to the boundary remaining number R _Lc , and the boundary remaining number C _Ln is cleared.

  The boundary comparison unit 63 compares the new boundary with the continuation boundary, and generates selection information for selecting boundary data having a small Y coordinate or boundary data having a small X coordinate when the Y coordinates are the same. When a new boundary exists and no continuation boundary exists, selection information for selecting a new boundary is generated. On the other hand, when a new boundary does not exist and a continuation boundary exists, selection information for selecting a continuation boundary is generated.

  The boundary selection unit 64 selects a new boundary or a continuation boundary as a subsequent boundary according to the selection information input from the boundary comparison unit 63, and stores the selected subsequent boundary in the subsequent boundary storage unit 65. The processing area storage unit 66 stores the processing area acquired from the subsequent boundary. The processing area is a lossless compression area of the line of interest, and is an area for decoding pixel data.

  Based on the subsequent boundary stored in the subsequent boundary storage unit 65 and the processing area stored in the processing area storage unit 66, the pixel number calculation unit 67 calculates the pixel number N1 of the lossless compression area and the pixel of the lossy compression area in the line of interest. Calculate the number N2.

  The pixel selection control unit 69 generates pixel selection information for processing pixels according to the pixel numbers N1 and N2 calculated by the pixel number calculation unit 67. The region color extraction unit 70 outputs the boundary color indicated by the subsequent boundary stored in the subsequent boundary storage unit 65 to the pixel selection unit 72 as pixel data.

  The pixel reading unit 71 reads the pixel data output by the lossy decoding unit 52 based on the pixel selection information, and outputs the read pixel data to the pixel selection unit 72. The pixel reading unit 71 asserts the ready signal via the above-described interface to notify the lossy decoding unit 52 that the pixel is in the ready state in order to read the pixel from the buffer 56 of the lossy decoding unit 52 based on the pixel selection information. . When both the valid signal and the ready signal are asserted, that is, when the lossy decoding unit 52 is in the valid state and the lossless decoding unit 53 is in the ready state, the pixel reading unit 71 reads the pixel data from the buffer 56. If both the valid signal and the ready signal are not asserted via the interface, the process waits until both the valid signal and the ready signal are asserted.

  The pixel selection unit 72 selectively outputs the pixel data output by the area color extraction unit 70 or the pixel data output by the pixel reading unit 71 to the writing processing unit 54 based on the pixel selection information. Therefore, the pixel selection unit 72 outputs, for each pixel, either pixel data output by the area color extraction unit 70 or pixel data output by the pixel reading unit 71. Here, the pixel data output by the pixel reading unit 71 in the pixel for which the pixel data output by the area color extraction unit 70 is selected is discarded.

Outline of Operation of Lossless Decoding Unit Although details will be described later, an outline of a processing procedure of the lossless decoding unit 53 will be described using the rectangular area 20 shown in FIG. 2 as an example. Each parameter of the lossless decoding unit 53 will be described with reference to FIG. Boundary reading unit 61, a boundary number of E = 5 is set to the boundary remaining number R _N, the boundary data of the first lossless compression region 21 (hereinafter, the boundary data 21) reads and stores the new boundary buffer. At this point, the continuation boundary queue 62 and the continuation boundary buffer are empty (state 0 in FIG. 14).

Since the continuous boundary buffer is empty, the boundary comparing unit 63 selects a new boundary as a subsequent boundary. As a result, the boundary data 21 is inputted to the FIFO of the continuation boundary queue 62, the boundary remaining number R _N is decremented (state 1). The decoding control unit 73 outputs 0xFF0000 as pixel data of three pixels from the left end of the first line (line 0) of the rectangular area 20 with reference to the boundary data 21.

  The boundary reading unit 61 reads boundary data of the second lossless compression area 23 (hereinafter, boundary data 23) and stores the boundary data in a new boundary buffer. On the other hand, the continuation boundary queue 62 stores the boundary data 21 extracted from the FIFO in the continuation boundary buffer (state 2).

The boundary comparison unit 63 compares the new boundary (boundary data 23) with the continuation boundary (boundary data 21), and selects a new boundary as a subsequent boundary. Consequently, FIFO boundary data 23 is input to the boundary remaining number R _N is decremented (state 3). The decoding control unit 73 outputs the pixel data read by the pixel reading unit 71 as pixel data of seven pixels from the fourth pixel of the first line (line 0) with reference to the boundary data 23. Further, 0x00FF00 is output as pixel data from the eleventh pixel to the right end, and the line of interest is incremented.

  The boundary reading unit 61 reads boundary data of the third lossless compression area 24 (hereinafter, boundary data 24) and stores the boundary data in a new boundary buffer. On the other hand, the boundary data 21 is still stored in the continuous boundary buffer (state 4).

  The boundary comparison unit 63 compares the new boundary (boundary data 24) with the continuation boundary (boundary data 21), and selects a continuation boundary as a subsequent boundary. As a result, the boundary data 21 is input to the FIFO (state 5). The decoding control unit 73 outputs 0xFF0000 as pixel data of six pixels from the left end of the line of interest (line 1) with reference to the boundary data 21.

  The continuation boundary queue 62 stores the boundary data 23 in the continuation boundary buffer. On the other hand, the new boundary buffer keeps storing the boundary data 24 (state 6).

  The boundary comparing unit 63 compares the new boundary (boundary data 24) with the continuation boundary (boundary data 23), and selects a continuation boundary as a subsequent boundary. As a result, the boundary data 23 is input to the FIFO (state 7). The decoding control unit 73 outputs the pixel data read by the pixel reading unit 71 as pixel data of three pixels from the seventh pixel of the line of interest (line 1) with reference to the boundary data 23. Further, 0x00FF00 is output as pixel data from the tenth pixel to the right end, and the line of interest is incremented.

  The continuation boundary queue 62 stores the boundary data 21 in the continuation boundary buffer. On the other hand, the new boundary buffer keeps storing the boundary data 24 (state 8).

  The boundary comparison unit 63 compares the new boundary (boundary data 24) with the continuation boundary (boundary data 21), and selects a continuation boundary as a subsequent boundary. As a result, the boundary data 21 is input to the FIFO (state 9). The decoding control unit 73 outputs 0xFF0000 as pixel data of six pixels from the left end of the line of interest (line 2) with reference to the boundary data 21.

  The continuation boundary queue 62 stores the boundary data 23 in the continuation boundary buffer. On the other hand, the new boundary buffer keeps storing the boundary data 24 (state 10).

  The boundary comparing unit 63 compares the new boundary (boundary data 24) with the continuous boundary (boundary data 23), and selects a new boundary as a subsequent boundary. As a result, the boundary data 24 is input to the FIFO (state 11). The decoding control unit 73 outputs the pixel data read by the pixel reading unit 71 as pixel data of two pixels from the eighth pixel of the line of interest (line 2) with reference to the boundary data 24. Further, 0x0000FF is output as pixel data of two pixels from the ninth pixel.

Boundary reading unit 61 decrements the border remaining number R _N, the boundary data of the fourth lossless compression region 26 (hereinafter, the boundary data 26) reads and stores the new boundary buffer. On the other hand, the boundary data 23 is still stored in the continuous boundary buffer (state 12).

  The boundary comparing unit 63 compares the new boundary (boundary data 26) with the continuation boundary (boundary data 23), and selects a continuation boundary as a subsequent boundary. As a result, the boundary data 23 is input to the FIFO of the continuous boundary queue 62 (state 13). The decoding control unit 73 outputs 0x00FF00 as pixel data from the twelfth pixel to the right end of the line of interest (line 2) with reference to the boundary data 23, and increments the line of interest.

The continuation boundary queue 62 stores the boundary data 21 in the continuation boundary buffer. On the other hand, the new boundary buffer keeps storing the boundary data 26 (state 14). Although the following description is omitted, the above processing procedure is repeated up to the last line of the rectangular area 20, and decoding of the rectangular area 20 is completed. In the above description, the description of the change in the remaining boundary numbers R _Lc and R _Ln is omitted, but the details will be described later.

[Decoding Process] The output process of the pixel data including the decoding process of the lossless decoding unit 53 will be described with reference to the flowchart of FIG. Note that the processing shown in FIG. 7 indicates processing for one rectangular area, and is repeatedly executed for the number of rectangular areas. The decoding control unit 73 initializes the lossless decoding unit 53 (S61). The initialization, the storage of the address of the source bit map data, setting the acquisition and border remaining number R _N boundary number E by Boundary reading unit 61, the boundary remaining number R _Lc continuous boundary queue 62, clearing of R _Ln, attention Initialization of the X coordinate of the line is included.

  Next, the decoding control unit 73 determines whether or not the line of interest is the last line of the rectangular area (S62). If the line of interest is the last line, the decoding processing of the rectangular area ends. When the line of interest is not the last line, the decoding control unit 73 uses the boundary reading unit 61, the continuation boundary queue 62, and the boundary determination unit 63 to execute acquisition of a subsequent boundary (S63) described later.

  Next, the decoding control unit 73 determines whether or not the acquired subsequent boundary corresponds to the first boundary of the line of interest (S64). If the subsequent boundary is the first boundary, the left side of the subsequent boundary is a lossy compression area. In this case, the decoding control unit 73 uses the pixel number calculation unit 67 and the pixel selection unit 72 to process the left lossy compression area (S65).

  The processing of the left-side lossy compression area calculates the number of pixels N2 of the lossy compression area based on the X coordinate of the start boundary of the subsequent boundary on the line of interest, and calculates the number of pixels N2 of the pixel data read by the pixel reading unit 71. This is a process of outputting the pixel data of (1) to the writing processing unit 54. In other words, pixel data from the left end of the line of interest to immediately before the start boundary is output. However, when the X coordinate of the start boundary of the first boundary is 0, there is no lossy compression area to the left of the lossless compression area, but since the number of pixels N2 = 0, the processing of step S65 is substantially performed. Absent.

  Next, the decoding control unit 73 determines a line end flag described later (S66), and if the line end flag is true, updates the processing area and the continuation boundary queue when the line is ended (S67). If the line end flag is false, the processing area and the continuation boundary queue when the line is not finished are updated (S68). Subsequently, the decoding control unit 73 performs decoding of the lossless compressed area (S69). Details of the processing in steps S67, S68, and S69 will be described later.

Next, the decoding control unit 73 determines whether or not the processing area has reached the right end of the line of interest (S70). If the processing area has not reached the right end of the line of interest, it is determined whether the remaining number of boundaries R _{Lc of the} line of interest is _greater than 0 (S71). If the processing area does not reach the right end of the line of interest and the remaining number of boundaries R _Lc > 0, the lossless compression area exists on the right side of the processing area, and the decoding control unit 73 returns the process to step S63.

When the processing area does not reach the right end of the line of interest and the remaining number of boundaries R _Lc = 0, only the lossy compression area exists on the right side of the processing area. In this case, the decoding control unit 73 uses the pixel number calculation unit 67 and the pixel selection unit 72 to process the right lossy compression area (S72).

  The processing of the right-side lossy compression area calculates the number of pixels N2 of the lossy compression area based on the X coordinate of the end boundary of the subsequent boundary on the line of interest, and calculates the number of pixels N2 of the pixel data read by the pixel reading unit 71. This is a process of outputting the pixel data of (1) to the writing processing unit 54. In other words, pixel data from immediately after the end boundary of the line of interest to the right end is output.

  When the processing area has reached the right end of the line of interest or after performing the processing for the right lossy compression area, the decoding control unit 73 returns the processing to step S62.

The following boundary acquisition process (S63) will be described with reference to the flowchart of FIG. The decoding control unit 73 initializes the line end flag to false (S81). Next, the boundary reading unit 61 determines the state of the boundary remaining number R _N and the new boundary buffer to determine the need to obtain the boundary data (S82). If the remaining boundary number R _N > 0 and the new boundary buffer is empty, the boundary data is acquired from the memory 55, and the acquired boundary data is stored in the new boundary buffer (S83).

On the other hand, the continuation boundary queue 62 determines the remaining number _RLc of the boundary of the line of interest and the state of the continuation boundary buffer in order to determine the necessity of setting the continuation boundary data (S84). If the remaining number of boundaries R _Lc > 0 and the continuous boundary buffer is empty, the boundary data extracted from the FIFO is stored in the continuous boundary buffer (S85).

Next, the boundary comparing unit 63 controls the boundary selecting unit 64 to select a subsequent boundary (S86). If the subsequent boundary is equal to the new boundary, a new boundary is selected as the subsequent boundary, and the new boundary is stored in the subsequent boundary storage unit (S87). “Subsequent boundary = new boundary” means that the new boundary buffer is not empty and the continuous boundary buffer is empty, or that the X coordinate of the new boundary is smaller than the X coordinate of the continuous boundary in the line of interest. In this case, the boundary reading unit 61 decrements the border remaining number R _N, emptying the new boundary buffer (S88).

If the subsequent boundary is equal to the continuation boundary, the continuation boundary is selected as the subsequent boundary, and the continuation boundary is stored in the subsequent boundary storage unit (S89). “Subsequent boundary = continuous boundary” means that the continuous boundary buffer is not empty and the new boundary buffer is empty, or that the X coordinate of the continuous boundary is smaller than the X coordinate of the new boundary in the line of interest. In this case, the continuation boundary queue 62 decrements the remaining line boundary number R _Lc of the line of interest, and empties the continuation boundary buffer (S90).

  When the new boundary buffer is empty and the continuous boundary buffer is empty, the boundary comparison unit 63 empties the subsequent boundary storage unit 65 to indicate that there is no subsequent boundary (S91).

  When “subsequent boundary = new boundary”, the decoding control unit 73 determines the start Y coordinate of the new boundary and the Y coordinate of the line of interest (S92). If they match, a subsequent boundary exists in the line of interest, and the process of acquiring the subsequent boundary ends.

If “subsequent boundary = continuous boundary”, the decoding control unit 73 determines the number of remaining boundaries R _Lc (S93). When the remaining number of boundaries R _Lc is 1 or more (R _Lc > 0), an unprocessed boundary exists on the target line. In this case, the decoding control unit 73 ends the subsequent boundary acquisition processing. When the remaining number of boundaries R _Lc is 0 (R _Lc = 0), the subsequent boundary is the last boundary on the line of interest.

In any of the following cases, there is no subsequent boundary in the line of interest, and the decoding control unit 73 sets the line end flag to true (S94), and ends the subsequent boundary acquisition processing.
When the subsequent boundary storage unit 65 is empty (S91),
-If the start Y coordinate of the new boundary ≠ the Y coordinate of the line of interest (S92),
_{-When the} remaining number of boundaries R _Lc = 0 of the line of interest is (S93).
Update of Processing Area and Continuation Boundary Queue When Line is Not Completed The update processing (S68) of the processing area and continuation boundary queue 62 when the line is not completed will be described with reference to the flowchart of FIG. The decoding control unit 73 sets the X coordinate of the end boundary from the X coordinate of the start boundary of the subsequent boundary of the target line as the updated processing area (S101).

  Next, the decoding control unit 73 determines whether the right end of the pre-update processing area stored in the processing area storage unit 66 and the left end of the post-update processing area are continuous (S102). If the two are not continuous, there is a lossy compression area between the processing area before update and the processing area after update. In this case, the decoding control unit 73 performs processing of the intermediate lossy compression area using the pixel number calculation unit 67, the pixel selection control unit 69, the pixel reading unit 71, and the pixel selection unit 72 (S103).

  In the processing of the intermediate lossy compression area, the number of pixels N2 of the lossy compression area is calculated based on the X coordinate of the processing area before and after the update on the target line, and the pixel data read by the pixel reading unit 71 is equivalent to the number of pixels N2. This is a process of outputting the pixel data of (1) to the writing processing unit 54. In other words, pixel data between the processing area before the update and the processing area after the update is output.

  Next, the decoding control unit 73 stores the updated processing area in the processing area storage unit 66 (S104), and determines whether the end Y coordinate of the subsequent boundary matches the Y coordinate of the line of interest (S105). ).

When the trailing Y coordinate of the succeeding boundary matches the Y coordinate of the line of interest, the decoding of the succeeding boundary is completed by decoding the next lossless compressed area (S69). No need to do. On the other hand, if the end Y coordinate of the subsequent boundary does not match the Y coordinate of the line of interest, the subsequent boundary needs to be held in the continuous boundary queue 62. Therefore, when the end Y coordinate of the subsequent boundary does not match the Y coordinate of the line of interest, the decoding control unit 73 inputs the subsequent boundary to the FIFO of the continuous boundary queue 62 and increments the remaining number R _Ln of boundary of the next line ( S106).

  Thus, the processing for updating the processing area and the continuation boundary queue 62 when the line is not completed is completed.

Update of Processing Area and Continuation Boundary Queue When Line Ends The update processing (S67) of the processing area and continuation boundary queue 62 when the line ends will be described with reference to the flowchart of FIG. The decoding control unit 73 increments the Y coordinate of the line of interest (S111), copies the remaining number R Ln of the next line of the continuous boundary queue 62 to the remaining number R Lc of the line of interest (S112), and The number R Ln is cleared (S113).

  Next, the decoding control unit 73 determines whether a new boundary is selected as a subsequent boundary (S114), determines whether a continuous boundary is selected (S115), and branches the process based on the determination result. .

  In the case of “subsequent boundary = new boundary”, there is no lossless compression area from the target line to the line before the leading Y coordinate of the new boundary. In this case, the decoding control unit 73 sets the leading Y coordinate of the subsequent boundary as the updated Y coordinate (S116). Then, while the condition (Y coordinate of target line <updated Y coordinate) is satisfied, the process of the lossy compression area of the line and the increment of the target line are repeated (S117).

  In the process of the line lossy compression area, for the line satisfying the condition, the width of the rectangular area is set to the pixel number N2 of the lossy compression area, and the pixel data of the pixel number N2 of the pixel data read by the pixel reading unit 71 is written. This is a process of outputting to the output processing unit 54. In other words, pixel data of the entire line is output.

  Next, the decoding control unit 73 sets the X coordinate of the end boundary from the X coordinate of the start boundary of the subsequent boundary to the updated processing area in the target line, and stores the updated processing area in the processing area storage unit 66. (S118), and the process proceeds to step S123.

  In the case of “subsequent boundary = continuous boundary”, the decoding control unit 73 sets the X coordinate of the right end of the rectangular area from the X coordinate of the start boundary of the subsequent boundary of the target line in the updated processing area, and The processing area is stored in the processing area storage unit 66 (S119), and the process proceeds to step S123.

  Further, when neither a new boundary nor a continuation boundary is selected as the subsequent boundary, in other words, when there is no subsequent boundary, there is no lossless compression area from the target line to the last line of the rectangular area. In this case, the decoding control unit 73 sets the Y coordinate of the last line as the updated Y coordinate (S120). Then, as long as the condition (Y coordinate of the line of interest ≦ updated Y coordinate) is satisfied, the process of the lossy compression area of the line and the increment of the line of interest are executed as in step S117. Subsequently, the decoding control unit 73 empties the processing area storage unit 66 (S122), and ends the processing for updating the processing area and the continuous boundary queue 62.

When there is a subsequent boundary, the decoding control unit 73 inputs the subsequent boundary to the FIFO of the continuous boundary queue 62, increments the remaining number R _Ln of the boundary of the next line (S123), and outputs the processing area and the continuous boundary queue 62. The update processing ends.

[Decoding of Lossless Compressed Area] The decoding processing (S69) of the lossless compressed area will be described with reference to the flowchart of FIG. The decoding control unit 73 determines whether or not the processing area storage unit 66 is empty (S131). If the processing area storage unit 66 is empty (the processing area is not set), the decoding processing of the lossless compression area is performed. To end.

  When the processing region is stored in the processing region storage unit 66, the decoding control unit 73 uses the pixel number calculation unit 67, the pixel selection control unit 69, the region color extraction unit 70, and the pixel selection unit 72 to set the processing region. Then, the pixel data of the area color indicated by the subsequent boundary is output to the writing processing unit 54 (S132). Further, the decoding control unit 73 deletes the pixel data input by the pixel reading unit 71 in the processing area (S133). As a result, the pixel data of the processing area (lossless compression area) of the target line is decoded.

[Image decoding operation]
Hereinafter, an image decoding operation in a system including the above-described image decoding device will be described.
● Configuration of Bitmap Data FIG. 12 shows an example of bitmap data. FIG. 12A shows a memory image of bitmap data. FIG. 12A shows that the bitmap data includes one bitmap 1, and the bitmap 1 has two rectangular areas in the horizontal and vertical directions (rectangular area 1, rectangular area 2, rectangular area 3, and rectangular area 4). 9 shows an example of bitmap data in the case of including. That is, the header of the bitmap 1 is stored, and lossless compressed data of the rectangular area and lossy compressed data of the rectangular area are stored after the header.

  FIG. 12B shows a configuration example of the header of bitmap 1. As indicated by the “bitmap size” field, the bitmap 1 is composed of a total of four rectangular areas, two in the horizontal direction W and two in the vertical direction H. As indicated by the “number of channels” field, bitmap 1 has three RGB channels. As indicated by the "bit precision" field, each channel is 8 bits.

  The pointer map of the lossless compressed data indicates the head address of the lossless compressed data of the rectangular area 1,..., The head address of the lossless compressed data of the rectangular area 4 in order from the top. The pointer map of the lossy compressed data indicates, in order from the top, the start address of the lossy compressed data of the rectangular area 1,..., The start address of the lossy compressed data of the rectangular area 4.

Configuration of Image Processing Device FIG. 13 is a block diagram showing an example of a system configuration of an image processing device including the above-described image decoding device.

  The CPU 141 controls the entire system by executing the OS and various programs stored in the storage unit 143 including the ROM and the flash memory using the RAM 142 as a work memory. The external interface (external I / F) 144 is a communication interface such as a USB card for connecting to an external device (for example, an information processing device or a scanner) or an Ethernet (registered trademark) card for connecting to a network. .

  The image decoding device 145 is the above-described image decoding device, and performs a decoding process on the encoded and compressed image data input via the external I / F 144. The image processing device 146 performs image processing such as performing high-quality processing on the image data decoded by the image decoding device 145 or converting the image data into the input data format of the printing device 147, for example. The printing device 147 forms an image on a recording medium using a recording material (toner or ink) based on the input image data.

  The interconnect 148 is a data transfer path between the CPU 141, the RAM 142, the storage unit 143, the external I / F 144, the image decoding device 145, the image processing device 146, and the printing device 147, which is configured by a shared bus, a switch, and the like.

  The external I / F 144 receives source bitmap data from an information processing device or a print server (not shown) via a serial bus or a network. The external I / F 144 notifies the CPU 141 that the data has been received by an interrupt signal or the like. The CPU 141 controls the external I / F 144 to transfer the source bitmap data to the RAM 142. FIG. 12A shows a memory image of the source bitmap data stored in the RAM 142.

  The CPU 141 notifies the image decoding device 145 of the start address 0x3000_0000 of the source bitmap data and the decoding instruction. As a method of notifying the decoding instruction, there is known a method in which the image decoding device 145 includes a control register for receiving the address of the source bitmap data and the decoding instruction, and the CPU 141 writes the address of the source bitmap data and the decoding instruction to the control register. Have been. Further, the CPU 141 notifies the image decoding device 145 of the head address of the buffer in which the decoded data is to be written.

  The control unit 51 (shown in FIG. 5) of the image decoding device 145 that has received the decoding instruction reads the header of the bitmap 1 from the address 0x3000_0000, and based on the bitmap size (W, H), the presence of four rectangular areas. Recognize. Then, the control unit 51 notifies the lossy decoding unit 52, the lossless decoding unit 53, and the writing processing unit 54 of the number 4 of rectangular areas, the number of channels 3, and the bit precision 8. With this notification, the lossy decoding unit 52, the lossless decoding unit 53, and the writing processing unit 54 are set to transfer the decoding result as 3 channels and 8 bits for each channel.

  Next, the control unit 51 reads the pointer of the lossless data of the rectangular area 1 from the pointer map of the lossless compressed data shown in FIG. 2B, and transfers the pointer to the lossless decoding unit 53. Further, the pointer of the lossy data of the rectangular area 1 is read from the pointer map of the lossy compressed data, and the pointer is transferred to the lossy decoding unit 52.

  The lossy decoding unit 52 reads the lossy data from the address 0x3000_0410 indicated by the pointer, decodes the data, and transfers the decoded pixel data to the lossless decoding unit 53 in line order (ascending order of Y coordinate) and in ascending order of X coordinate. I do.

  The lossless decoding unit 53 reads the lossless data from the address 0x3000_0010 indicated by the pointer, decodes the data, and transfers the decoded pixel data to the writing processing unit 54 in line order (ascending order of Y coordinate) and in ascending order of X coordinate. I do.

  As described above, in a rectangular area in which the lossy compression area and the lossless compression area are mixed with an arbitrary boundary interposed therebetween, the expansion result of the lossy compression data or the expansion result of the lossless compression data is output in raster order while referring to the boundary information. Therefore, all of the expansion results of the lossy compressed data are stored in the memory, and the encoded data of the rectangular area can be obtained without overwriting the lossless compressed area with the expansion results of the lossless compressed data. In other words, the pixel data transferred to the RAM 142 via the interconnect 148 is pixel data of the lossy compression area decoded from the lossy compression data, and / or pixel data of the lossless compression area decoded from the lossless compression data. . That is, it is not necessary to transfer pixel data in the lossless compression area decoded from the lossy compression data via the interconnect 148, so that unnecessary consumption of the system bus band and the memory band can be prevented.

[Modification]
A modified example of the image decoding device is shown by the block diagram of FIG. The image decoding device illustrated in FIG. 15 includes a coordinate management unit 74 instead of the pixel number calculation unit 67. The coordinate management unit 74 manages the coordinates of the pixel being decoded in the bitmap data, and determines whether the coordinates of the pixel belong to a lossless compression area with the start boundary at the right end or a lossy compression area with the end boundary at the right end. Is generated. That is, the determination process is performed for each pixel.

  The coordinate management unit 74 sequentially counts from the origin coordinates (0, 0) to the coordinates (W-1, H-1) of the bitmap data to be decoded (the horizontal and vertical pixel numbers are H and W, respectively). Up. That is, (0, 0), ..., (W-1, 0), (0, 1), ..., (W-1, 1), ..., (0, H-1), ..., (W-1, As shown in H-1), the coordinates are counted up in the main scanning line direction, and when reaching the right end, the coordinates are counted up in the sub scanning line direction. That is, the coordinate management unit 74 generates data indicating the coordinates in the raster scan order. Then, the coordinate management unit 74 generates control information for always processing one pixel for the pixel selection control unit 69.

[Second embodiment]
Hereinafter, an image decoding apparatus and an image decoding method according to the second embodiment of the present invention will be described. In the second embodiment, the same components as those in the first embodiment are denoted by the same reference numerals, and detailed description thereof may be omitted.

  An example of the configuration of an image processing apparatus having a rendering system to which the image decoding apparatus according to the second embodiment is applied is shown in the block diagram of FIG. Note that the same components as those shown in FIG. 13 are denoted by the same reference numerals, and detailed description thereof will be omitted.

  The rendering system shown in FIG. 16 converts print data described in a page description language (PDL) into intermediate data, performs rendering based on the intermediate data, and generates bitmap data. The print data includes one or more drawing elements, and the drawing elements include text, graphic figures, and bitmap images.

  The intermediate data generation unit 151 is implemented as software executed by the CPU 141, converts print data read from the RAM 142 into intermediate data, and writes the intermediate data into the RAM 142. The generated intermediate data includes renderer control information, boundary information of each drawing element, paint information and combination (blend) information, and a bitmap image. In addition, the bitmap image included in the intermediate data is often encoded and compressed.

  The renderer control unit 152 controls the renderer 155 based on the renderer control information included in the intermediate data. The renderer control unit 152 includes the function of the control unit 51 of the image decoding device, and outputs control information to the rendering element processing unit 153 and the renderer writing processing unit 154.

  The drawing element processing unit 153 reads the boundary information, the painting information, and the combination information of the drawing element included in the intermediate data, generates pixel data forming a page, and outputs the pixel data to the renderer writing processing unit 154. When the drawing element is a bitmap image, the drawing element processing unit 153 reads the bitmap image from the RAM 142 based on the control information, and synthesizes a graphic figure or text pixel with the bitmap image subjected to the interpolation processing as necessary. I do.

  The renderer writing processing unit 154 writes the pixel data input from the drawing element processing unit 153 to a predetermined address of the RAM 142.

  The renderer 155 includes an image decoding unit and an intermediate data processing unit. The image decoding unit includes a renderer control unit 152, a lossy decoding unit 52, a lossless decoding unit 53, and a writing processing unit 54. The intermediate data processing unit includes a renderer control unit 152, a rendering element processing unit 153, and a renderer writing processing unit 154. As described above, the renderer control unit 152 is a component of the image decoding unit and the intermediate data processing unit.

  When the intermediate data includes the encoded and compressed image data, the renderer control unit 152 outputs the decoding control information of the compression-encoded image data to the image decoding unit according to the renderer control information of the intermediate data. The image decoding unit decodes the bitmap image encoded and compressed as described in the first embodiment, and writes the decoding result into the RAM 142.

  Next, the renderer control unit 152 outputs control information to the intermediate data processing unit according to the renderer control information of the intermediate data. The drawing element processing unit 153 reads the bitmap image decoded by the image decoding unit and written in the RAM 142, and combines the bitmap image with pixel data of another drawing element.

  The drawing element processing unit 153 outputs each pixel data obtained by the synthesis processing to the renderer writing processing unit 154. The renderer writing processing unit 154 writes pixel data to a predetermined address of the RAM 142. By the above rendering processing, the print data is converted into bitmap data and stored in the RAM 142.

[Other Examples]
The present invention supplies a program realizing one or more functions of the above-described embodiments to a system or an apparatus via a network or a storage medium, and one or more processors in a computer of the system or the apparatus read and execute the program. Processing can also be realized. It can also be realized by a circuit (for example, an ASIC) that realizes one or more functions.

  Note that the lossy decoding unit 52 can also be realized by software processing in which a processor provided separately from the CPU and having a dedicated work memory reads and executes a program from a dedicated program memory or a shared system memory. The lossless decoding unit 53 can also be realized by a circuit that realizes the same function. Further, both the lossy decoding unit 52 and the lossless decoding unit 53 may be processors, and both the lossy decoding unit 52 and the lossless decoding unit 53 may be dedicated circuits.

52: Lossy decoding unit, 53: Lossless decoding unit

Claims (20)

First data obtained by compressing an image by a first method for each area and second data obtained by compressing a partial area of the image by a second method different from the first method are mixed. An image decoding device that decodes data and outputs the image,
First decoding means for decoding the first data and outputting pixel data of each pixel constituting an image;
Second decoding means for decoding the second data in the compressed area based on boundary information indicating a compressed area compressed by the second method;
Writing means for writing pixel data of each pixel constituting the image obtained from the second decoding means to a memory,
The first decoding means outputs pixel data of each pixel to the second decoding means,
The second decryption means includes:
Determine whether the processing target pixel is the compression area based on the boundary information,
Pixel data obtained from the first decoding unit is output to the writing unit for pixels other than the compressed region, and pixel data obtained by decoding by the second decoding unit for pixels included in the compressed region. Is output to the writing means.   2. The image decoding apparatus according to claim 1, wherein, of the pixel data obtained from the first decoding unit, pixel data of a pixel included in the compression area is not stored in the memory.   2. The image decoding apparatus according to claim 1, wherein the writing unit writes pixel data to the memory once for each pixel constituting the image. The second decryption means includes:
Acquisition means for acquiring the boundary information as new boundary information,
Holding means for holding the boundary information as continuation boundary information;
Selecting means for selecting one of the new boundary information and the continuation boundary information as subsequent boundary information indicating the succeeding compressed area;
Claims the subsequent boundary information input to said holding means, and having a control unit for decoding and selection of the pixel data of the second data of the compressed area based on the subsequent boundary information Item 4. The image decoding device according to any one of Items 1 to 3. The boundary information includes, for each line, start information indicating the left end of the compressed area and end information indicating the right end of the compressed area,
Wherein said control means includes a characterized in that it is carried out in each of the lines, the selection of decoding and the pixel data of the second data of the compressed area based on the start information and the end information the indicating subsequent boundary information The image decoding apparatus according to claim 4, wherein The boundary information further includes color information of the compression area,
The control means, as a decoding of the second data of the compressed region, from said start information of the pixel shown coordinates to the coordinates of the pixel indicated by the end information, to output the color information as the pixel data The image decoding apparatus according to claim 5, wherein: The holding unit includes a first counter for managing the number of inputs of the subsequent boundary information, and a second counter for managing the number of continuation boundary information held for a line of interest. The image decoding device according to any one of claims 4 to 6, wherein The image according to claim 7, wherein the holding unit outputs the continuation boundary information in the order of input, and decrements the second counter when the continuation boundary information is selected as the subsequent boundary information. Decoding device. If the new boundary information is selected as the subsequent boundary information, and if the first line of the compressed area indicated by the new boundary information does not coincide with the line of interest, the continuation boundary information is selected as the subsequent boundary information. In the second case where the second counter indicates 0, or in the third case where there is no boundary information to be selected as the subsequent boundary information, the holding unit stores the information of the first counter in the second The image decoding apparatus according to claim 8, wherein the first counter is cleared by copying to the counter, and the control unit increments the coordinates of the line of interest. In the first case, while the coordinates of the line of interest do not reach the head line of the compression area indicated by the subsequent boundary information, the control unit may increment the coordinates of the line of interest and determine whether the first decoding unit The image decoding apparatus according to claim 9, wherein selection of decoded pixel data is repeated. In the third case, the control unit increments the coordinates of the line of interest and selects the pixel data decoded by the first decoding unit until the coordinates of the line of interest reaches the coordinates of the last line of the image. 11. The image decoding device according to claim 9, wherein The acquisition unit acquires the number of the compression region included in the image, a third counter for managing the remaining number of the compression region, the new boundary information is selected as the subsequent boundary information The image decoding device according to any one of claims 4 to 11, wherein the third counter is decremented when done. The image decoding apparatus according to claim 12, wherein the acquisition unit acquires the new boundary information when the third counter indicates the presence of the compression area after the decrement.   14. The image decoding apparatus according to claim 1, wherein the first method is a lossy compression method, and the second method is a lossless compression method. Said second decoding means, the image decoding apparatus according to claim 4, based on two of the boundary information to be contiguous, and performs selection of the first data. Wherein in the second region for selecting the data, the image decoding apparatus according to claim 15, wherein the first data is discarded. The first decoding means is constituted by dedicated hardware,
The second decoding unit is configured by a processor having a dedicated work memory, and the processor executes a program for causing the processor to function as the second decoding unit. 2. The image decoding device according to claim 1.   An image processing device that performs rendering using the image decoding device according to claim 1. First data obtained by compressing an image by a first method for each area and second data obtained by compressing a partial area of the image by a second method different from the first method are mixed. A method for controlling an image decoding device that decodes data and outputs the image.
First decoding means for decoding the first data, outputting pixel data of each pixel constituting an image,
A second decoding unit that decodes the second data of the compression area based on boundary information indicating a compression area compressed by the second method;
Writing means writes pixel data of each pixel constituting the image obtained by decoding the second data to a memory,
Pixel data of each pixel obtained by decoding the first data is passed to a decoding process of the second data,
In the decoding of the second data,
Determine whether the processing target pixel is the compression area based on the boundary information,
For pixels other than the compressed region, pixel data obtained by decoding the first data is output as the writing target, and for pixels included in the compressed region, pixels obtained by decoding the second data An image decoding method, wherein data is output as the writing target. A program for causing a computer to function as each unit of the image decoding device according to any one of claims 1 to 17 .

Images