JP2017212601A - Image processing device, image processing method, program, and data - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the performance by reducing a cache mistake at the time of performing processing of an input image in such a configuration that the input image formed of pixels having a color value and an attribute value is encoded, and the encoded input image is decoded for use at the time of processing of the input image.SOLUTION: An image processing device identifies a boundary where a color value changes and a boundary where an attribute value changes in an input image formed of pixels having the color value and the attribute value, generates and outputs data containing a first region storing information indicating the color value and the attribute value of a region partitioned by the boundary and a second region storing the position information of the boundary.SELECTED DRAWING: Figure 3

Description

  The present invention relates to an image processing apparatus, an image processing method, a program, and data.

  An image processing technique is known in which, when an image is processed, an attribute map representing pixel attributes is referred to and processing is changed according to the attributes. For example, image area separation processing is performed on an image obtained by scanning, attribute map information indicating whether each pixel is a character area pixel or a photographic area pixel is generated, and image processing is performed according to the attribute map information. Parameters can be set. For example, it is possible to obtain a good output image by sharpening the character part by performing edge enhancement processing on the pixels in the character area and smoothing processing on the pixels in the photographic area. Become.

  Further, Patent Document 1 discloses the following method in order to prevent image quality deterioration in encoding of mixed images including pixels having different attributes. First, a foreground image area including a character / line image pixel is determined, binary image data is generated from multivalued image data in the foreground image area, and lossless compression is performed. Second, the value of the multi-value pixel at the position of the character / line image pixel is replaced with the value of the multi-value pixel at the position of the non-character / line image pixel, and the obtained image (background image) is irreversibly compressed. . Third, the attribute data for the pixels in the character / line drawing area is masked with a preset value, and the attribute data is losslessly compressed. With the above method, the image quality of the output image can be improved while reducing the data amount of the attribute data.

Japanese Patent No. 5132530

  However, in the method described in Patent Document 1, the foreground image, the background image, and the attribute data are individually compressed. For this reason, at the time of decoding, the color data of the foreground image, the color data of the background image, and the attribute data are developed in separate memory areas. When processing is performed with reference to these data, it is necessary to read out the color data and the attribute data from the respective memory areas. Even when color data and attribute data are combined for each pixel before processing, it is still necessary to read color data and attribute data from each memory area. Then, if color data and attribute data are read from the respective memory areas, a cache miss is likely to occur without a hardware resource such as a secondary cache for efficiently supplying these data. There was a problem.

  The present invention encodes an input image composed of pixels having color values and attribute values, and decodes and uses the encoded input image when processing the input image. The purpose is to reduce cache misses and improve performance when processing.

In order to achieve the object of the present invention, for example, an image processing apparatus of the present invention comprises the following arrangement. That is,
A specifying means for specifying a boundary where a color value changes and a boundary where an attribute value changes in an input image including pixels having a color value and an attribute value;
First compression for generating and outputting data including a first area for storing information indicating color values and attribute values of an area delimited by the boundary, and a second area for storing position information of the boundary Means,
Is provided.

  When an input image composed of pixels having color values and attribute values is encoded, and when the input image is processed, the encoded input image is decoded and used. In addition, cache misses can be reduced and performance can be improved.

1 is a configuration diagram of an image processing system according to an embodiment. The figure explaining the concept of a flattening process. The block diagram of the encoding part which concerns on one Embodiment. The block diagram of the 2nd encoding part which concerns on one Embodiment. An example of a memory image of compressed bitmap data included in intermediate data. The figure which shows the example of the rectangular area | region in a bitmap, and the area | region contained therein. The figure which shows the example of the rectangular area | region in a bitmap, and the boundary of a lossless compression area | region. The figure which shows the example of the rectangular area | region in a bitmap, and the boundary of an irreversible compression area | region. The figure which shows an example of the format of lossless compression data. The block diagram of the 2nd decoding part and 1st output part which concern on one Embodiment. 5 is a flowchart of an image processing method according to an embodiment.

  Embodiments of the present invention will be described below with reference to the drawings. However, the scope of the present invention is not limited to the following embodiments.

  Image data composed of pixel groups having color values and attribute values when encoded image data is decoded by using an image processing technique according to an embodiment of the present invention described below. Is expanded in memory. For this reason, a pixel group having a color value and an attribute value is used without using a process for combining the color value and the attribute value for each pixel or an additional hardware resource for improving the efficiency of the process. It is possible to efficiently use the processed image data in the post-decoding process.

  The image processing technique according to the present invention is generally applicable to image encoding and decoding. For example, when encoding and decoding an image in a configuration in which a computer encodes an image, transmits the encoded image to a printer, decodes the transmitted image, performs image processing, and then performs printing. In addition, the image processing technique according to the present invention can be applied. Hereinafter, as an application example of the image processing technology according to the present invention, print data described using a page description language is transmitted to a printer, and the printer converts the print data into intermediate data and temporarily stores the data. explain. In this case, the printer generates a print image that is bitmap data by rendering from the temporarily stored intermediate data, and performs a print process based on the print image.

  Here, the intermediate data includes a plurality of drawing elements, and the print image is generated by synthesizing the drawing elements defined by the intermediate data. The intermediate data may include several bitmaps such as a bitmap transmitted as print data and a bitmap generated by flattening processing described later. Such a bitmap can be encoded and compressed in order to save a storage area for intermediate data. The image processing technique according to the present invention can be applied to such bitmap encoding and decoding, for example.

  In the following embodiment, a reversible compression area and an irreversible compression area are set in one image (for example, one bit map), and only the image in the lossless compression area is reversibly compressed. Will be described with respect to a case where lossless compression is not performed. However, the image processing technique according to the present invention can also be applied to a case where an entire image is compressed by a single method.

  FIG. 1 shows an example of the configuration of an image processing system 100 that executes an image compression method and an image expansion method according to the present embodiment. The image processing system 100 may be, for example, a single printer or a printing system including a plurality of devices. In FIG. 1, a CPU 101 controls the entire system. Further, the CPU 101 converts print data described in PDL into intermediate data that is interpreted by a renderer 105 described later. This process is called intermediate data generation process. The image compression method according to the present embodiment is applied to compress image data according to print data in order to generate intermediate data including compressed image data in the intermediate data generation process. In one embodiment, the CPU 101 compresses a bitmap included in the intermediate data.

  The RAM 102 includes, for example, a memory controller and a DRAM. A program can be expanded in the RAM 102, and the CPU 101 performs processing according to the program expanded in the RAM 102, whereby an operation described later can be realized. The RAM 102 can be used as a work memory for the encoding unit 301 that operates on the CPU 101 and the image processing system 100. Further, the RAM 102 can store print data, intermediate data, and print images.

  The ROM 103 stores a program. The CPU 101 can perform processing according to this program. An external interface (IF) 104 is an interface for connecting the image processing system 100 to an external device. By using a technology such as Ethernet or USB via the external IF 104, the image processing system 100 can transmit and receive data to and from an external device. The renderer 105 performs rendering processing to convert intermediate data into a print image.

  The image processing device 106 performs image processing on the print image. For example, the image processing apparatus 106 can perform high-quality image processing on a print image or convert the print image into a data format that can be input to the printing apparatus 107. When each pixel of the print image has an attribute value, high-accuracy image quality processing can be performed with reference to the attribute value. As an example, edge enhancement processing can be performed on pixels having character attributes, and smoothing processing can be performed on pixels having photo attributes. However, the image processing apparatus 106 may perform high image quality processing on the bitmap decoded by the renderer 105. The printing apparatus 107 forms toner images on the paper surface by fixing toner and ink on the paper surface based on the print image that is electronic information.

  The interconnect 108 is a data transfer path between the CPU 101, the external IF 104, the renderer 105, the image processing device 106, the printing device 107, the RAM 102, and the ROM 103. The interconnect 108 can be configured using a shared bus, a switch, or the like.

  The renderer 105 includes an image decoding device and an intermediate data processing device. The image decoding apparatus includes a control unit 109, a first decoding unit 110, a second decoding unit 111, and a first output unit 112. The intermediate data processing apparatus includes a control unit 109, a drawing unit 113, and a second output unit 114. In the example of FIG. 1, the control unit 109 is a component common to the image decoding device and the intermediate data processing device, but the image decoding device and the intermediate data processing device may each have a control unit.

  The control unit 109 reads renderer control information from the intermediate data, and controls the renderer 105 according to the renderer control information. Specifically, the control unit 109 transfers control information to the first decoding unit 110, the second decoding unit 111, the first output unit 112, the drawing unit 113, and the second output unit 114. Control each of these parts.

  The first decoding unit 110 reads and decodes the irreversible compressed data included in the intermediate data. By decoding the lossy compressed data, pixel data including color values for each pixel can be obtained. Then, the first decoding unit 110 transfers pixel data composed of color values obtained by decoding to the second decoding unit 111.

  The second decoding unit 111 reads and decodes the lossless compressed data included in the intermediate data. By decoding the lossy compressed data, pixel data including a color value and an attribute value can be obtained for each pixel. Then, the second decoding unit 111 transfers the pixel data obtained by the decoding and the pixel data transferred from the first decoding unit 111 to the first output unit 112. The first output unit 112 writes the transferred pixel data in a predetermined memory area such as the RAM 102. The first output unit 112 outputs the color value and attribute value decoded by the second decoding unit for the pixels in the lossless compression region, and the color decoded by the first decoding unit for the pixels in the lossy compression region. The value and the attribute value decoded by the second decoding unit can be output.

  The drawing unit 113 generates pixel data constituting a print image according to the intermediate data stored in the RAM 102. For example, the drawing unit 113 can generate a print image by a combining process of drawing elements defined by intermediate data. Specifically, the drawing unit 113 can read the boundary information, the paint information, and the synthesis information of the drawing elements from the intermediate data and synthesize overlapping drawing elements. The drawing unit 113 then transfers the generated pixel data to the second output unit 114. As an example of specific synthesis processing, for each pixel, the color value of the drawing element and the transmittance that is the attribute value are used as inputs, and the color values of a plurality of drawing elements that occupy the coordinates are synthesized according to a predetermined calculation formula. There is a way to do it. Such a color synthesis method is called α blend and is widely known. The second output unit 114 transfers the pixel data transferred from the drawing unit 113 to a predetermined address such as the RAM 102, for example.

  When the compressed data is included in the intermediate data, the control unit 109 transmits the decoded bitmap decoding control information to the first decoding unit 110 and the second decoding unit 111 according to the renderer control information included in the intermediate data. Forward. The first decoding unit 110 and the second decoding unit 111 decode the compressed bitmap according to the decoding control information, and transfer it to the RAM 102 via the first output unit 112.

  Next, the control unit 109 transfers the control information to the drawing unit 113 according to the renderer control information included in the intermediate data. The drawing unit 113 reads the bitmap decoded by the first decoding unit 110 and the second decoding unit 111 and transferred to the RAM 102 by the first output unit 112, and performs a synthesis process with other drawing elements. . Further, the pixel data obtained by the synthesis process is transferred to the second output unit 114. The second output unit 114 writes pixel data to a predetermined address in the RAM 102. Through the above rendering process, the intermediate data is converted into a print image and recorded in the RAM 102.

  As the bitmap included in the intermediate data, there are a bitmap included in the print data from the beginning and a bitmap generated by a partial rendering process called a flattening process.

  Here, the flattening process will be described. The intermediate data generation process executed by the CPU 101 operates using a memory area having a specified size. On the other hand, in the intermediate data generation process, when processing an area that includes many drawing elements or a complicated drawing element, there may be a shortage of memory necessary for the process. Therefore, when the allocated memory is insufficient to perform the process for the predetermined area, the bitmap of this area is generated by performing the bitmap process and the α blend process for the plurality of drawing elements. This process is called flattening process.

  At the time of flattening, the attribute value corresponding to the bitmap is generated with reference to the attribute value of the drawing element to be flattened. The attribute value indicates an attribute for each pixel of the bitmap, and can include, for example, the type of drawing element corresponding to each pixel, the transmittance of each pixel, and the like. In the present embodiment, the attribute value is represented as, for example, an area attribute map indicating the type of drawing element corresponding to each pixel of the bitmap, a transmittance map indicating the transmittance α corresponding to each pixel of the bitmap, or the like. The In addition, the bitmap included in the intermediate data is often compressed and encoded. The attribute value generated by the flattening process is also compression encoded together with the corresponding bitmap. The attribute value may be generated by the image processing system 100 as described above, but may be included in the print data from the beginning together with the bitmap.

  FIG. 2 shows the concept of the flattening process. FIG. 2 shows a rectangular area 201 to be processed, a bitmap 202 with 0 transmittance, a circular graphic figure 203 with 0 transmittance, a triangular graphic figure 204 with 0.5 transmittance, and a text 205 with 0 transmittance. ing. A bitmap 206 is generated by the flattening process.

  Specifically, the bitmap 202, the circular graphic figure 203, the triangular graphic figure 204, and the text 205 are sequentially overlapped by flattening processing to generate a bitmap 206. At this time, the following region attribute map is also generated. That is, a background attribute is assigned to a pixel in a region where none of the drawing elements 202, 203, 204, and 205 exists. A bitmap image attribute is assigned to a pixel in an area where only the bitmap 202 exists. Graphic attributes are assigned to pixels of an area where only the graphic figure 203 exists, an area where only the graphic figure 204 exists, and an area where the graphic figure 203 and the graphic figure 204 overlap. Attributes determined according to a predetermined calculation are assigned to pixels in an area where the bitmap 202 and the graphic figure 204 overlap. Text attributes are assigned to pixels in the area where the bitmap 202 and text 205 overlap.

  Further, when the bitmap 206 is generated by flattening processing, the following transmittance map is also generated. That is, 1 is set as the transmittance for the pixels in the region where none of the drawing elements 202, 203, 204, and 205 exists. For the pixels in the area where only the graphic figure 204 exists, 0.5 is set as the transmittance. Since the other area is occupied by a drawing element having a transmittance of 0 or a drawing element including a drawing element having a transmittance of 0 is overlapped, 0 is set as the transmittance of pixels in these areas.

  In this way, the bitmap generated by flattening can be composed of a plurality of regions having different attribute values. When a print image is generated using a bitmap generated by flattening, both the color value and attribute value (transmittance) of the bitmap are required. As described above, the image processing technique according to the present embodiment can be applied to compression of a bitmap generated by flattening.

  In the following, encoding of a bitmap in which each pixel has a transparency as an attribute value will be mainly described. However, the attribute value is not limited to the transmittance, and may be, for example, area information indicating whether the pixel is a character area pixel or a photographic area pixel as described above. Each pixel may have a combination of two or more attribute values.

  FIG. 3 shows an example of the configuration of an encoding unit that implements the image compression method according to the present embodiment. The encoding unit 301 includes an area dividing unit 302, an image replacing unit 303, a first encoding unit 304, an attribute combining unit 305, and a second encoding unit 306. The functions of these units can be realized by, for example, expanding a program stored in the ROM 103 or the like in a memory such as the RAM 102 and operating a processor such as the CPU 101 according to the expanded program.

  The encoding unit 301 receives an input image composed of pixels having color values and attribute values. This input image is a bitmap included in the intermediate data, and can be, for example, a bitmap included in the print data, a bitmap generated by flattening processing, or the like. In the following description, it is assumed that the input image is an RGB 8-bit bitmap. In this input image, the color values (RGB values) of the respective pixels are recorded on the RAM 102 in the following order.

  R (0, 0), G (0, 0), B (0, 0), R (1, 0) G (1, 0) B (1, 0), ..., R (X-max, 0 ), G (X-max, 0), B (X-max, 0), R (0, 1), G (0, 1), B (0, 1), R (1, 1), G ( 1, 1), B (1, 1), ..., R (X-max, 1), G (X-max, 1), B (X-max, 1), ..., R (0, Y) -Max), G (0, Y-max), B (0, Y-max), R (1, Y-max), G (1, Y-max), B (1, Y-max), ... ..., R (X-max, Y-max), G (X-max, Y-max), B (X-max, Y-max)

  Here, R (X, Y), G (X, Y), and B (X, Y) indicate the RGB values of the pixels at the coordinates (X, Y) of the bitmap, respectively. Further, when the width or height of the input image is not a multiple of 16, for example, pixels having predetermined color values and attribute values are padded at the right end and the lower end of the image so as to be a multiple of 16. X-max is the X coordinate of the rightmost pixel of the input image, and is represented by 16m-1 (m is a natural number). Y-max is the Y coordinate of the lower end pixel of the input image, and is represented by 16n-1 (n is a natural number).

  In the following description, it is assumed that input attribute data indicating an attribute value is an 8-bit transmittance map for each pixel. On the RAM 102, the transmittance of each pixel is recorded in the following order. In the following, α (X, Y) indicates the transmittance α of the pixel at the coordinate (X, Y) of the bitmap.

  α (0,0), α (1,0),..., α (X−max, 0), α (0,1), α (1,1),. ), ..., α (0, Y-max), α (1, Y-max), ..., α (X-max, Y-max)

  The area dividing unit 302 sets a lossless compression area and a lossy compression area in the input image. Thus, the region dividing unit 302 divides the input image into an irreversible compression region and a lossless compression region. The area dividing unit 302 generates area specifying information for identifying whether each pixel of the input image is a lossy compression area or a lossless compression area.

  In this embodiment, the area dividing unit 302 sets an area composed of a predetermined number of pixels having the same color value and attribute value and adjacent to each other as one lossless compression area. In the present embodiment, the predetermined number is 3 pixels, but is not particularly limited. In this embodiment, one reversible compression area is configured by pixels having exactly the same color value and attribute value. However, depending on the application, the color value and attribute value of a pixel included in one reversible compression area are less than or equal to a threshold value. There may be errors. Further, when the lossless compression area and the lossy compression area are set in advance in the input image, or when the lossless compression area and the lossy compression area are not set, the area dividing unit 302 is unnecessary.

  Hereinafter, a specific processing example of the area dividing unit 302 will be described. The area dividing unit 302 reads bitmap pixel data from the RAM 102 in units of 16 × 16 pixel rectangles. Here, the area dividing unit 302 sequentially reads pixel data from the upper left pixel of the rectangular area in raster order, and buffers pixel data for three consecutive lines of the read pixel data. Then, the area dividing unit 302 refers to the buffered pixel data for three lines, and when the RGB values are the same for three or more adjacent pixels, the area dividing unit 302 is configured by a group of pixels having the same RGB value. Is determined to be a lossless compression region. Here, the fact that the pixels are adjacent means that the X coordinate difference is 0 or 1 and the Y coordinate difference is 0 or 1 for two pixels. On the other hand, if the RGB values of adjacent pixels are the same even if only two adjacent pixels have the same RGB value, or the RGB values of any adjacent pixels are different, An area composed of the pixels is determined as an irreversible compression area.

  Then, the area dividing unit 302 assigns area specifying information indicating the lossless compression area to the pixel determined to be the lossless compression area, and the pixel determined to be the lossy compression area is the lossy compression area. Allocate area specifying information indicating that there is. The area specifying information generated for each pixel is supplied to the image replacement unit 303 and the attribute combination unit 305 in the same order as the arrangement of the color values and attribute values of the pixels included in the bitmap.

  The image replacement unit 303 generates an irreversible compression target image by replacing the color value of the pixel in the lossless compression region with a predetermined value. In the present embodiment, the image replacement unit 303 generates an irreversible compression target image by replacing the RGB values of the pixels in the lossless compression region with reference to the region specifying information. In the present embodiment, the image replacement unit 303 replaces the RGB values of the pixels in the lossless compression area with the replacement values calculated with reference to the RGB values of the pixels in the lossy compression area. For example, the image replacement unit 303 can read input image data in a rectangular unit of 16 × 16 pixels and generate a lossy compression target image in a rectangular unit of 16 × 16 pixels. At this time, the image replacement unit 303 uses the average R value, the average G value, and the average B value of the pixels in the lossy compression region among the pixels in the rectangular region, and the R value of the pixels in the lossless compression region, G Each of the value and the B value can be replaced.

  As described above, the first encoding unit 304 performs lossy compression on the input image generated by replacing the color value of the pixel in the lossless compression region with a predetermined value. By this processing, it is possible to suppress image deterioration in the lossy compression region when the lossy compression target image is subjected to lossy compression and decoding. However, the replacement process by the image replacement unit 303 is not essential. Note that a first encoding unit 304, an attribute combining unit 305, and a second encoding unit 306, which will be described later, can perform processing in the same rectangular units as the replacement processing performed by the image replacement unit 303.

  The first encoding unit 304 generates lossy compressed data representing an input image in the lossy compression region. For example, the first encoding unit 304 can generate lossy compressed data by irreversibly compressing the lossy compression target image in a 16 × 16 pixel rectangular unit. Although the compression method is not particularly limited, for example, the JPEG method can be mentioned. In the present embodiment, the entire irreversible compression target image is compressed, but only the input image in the irreversible compression area may be compressed. In the present embodiment, the attribute value is not included in the lossy compressed data.

  The attribute combining unit 305 sequentially reads input image data and attribute data from the upper left pixel of the rectangular area in units of 16 × 16 pixels. The attribute combining unit 305 combines the RGB value of the input image data and the attribute data for each pixel. In the present embodiment, the attribute combining unit 305 combines RGB values as color values and α values as attribute values. Thus, the attribute combining unit 305 generates a lossless compression target image having RGBα values for each pixel. The data of the lossless compression target image is arranged in the rectangular order as described below, and is supplied to the second encoding unit 306.

  [Rectangle (0, 0)], [Rectangle (1, 0)], [Rectangle (2, 0)], ..., [Rectangle (X'max, 0)], [Rectangle (1, 0)], [Rectangle (1, 1)], ..., [Rectangle (X'max, Y'max-1)], [Rectangle (0, Y'max)], [Rectangle (1, Y'max)], ... ..., [Rectangle (X'max, Y'max)],

  Here, the rectangle (X, Y) indicates the Xth rectangle from the left and the Yth rectangle from the top. X′max and Y′max represent the number of rectangles in the horizontal and vertical directions, respectively, and X′max = X−max−15) / 16, Y′max = (Y−max−15) / 16) It is. Each rectangular image data is arranged in pixel order as follows. In the following, R (X, Y), G (X, Y), B (X, Y), and α (X, Y) are the R values of the Xth pixel from the left and the Yth pixel from the top in the rectangle, respectively. , G value, B value, and α value.

  R (0,0), G (0,0), B (0,0), α (0,0), R (1,0), G (1,0), B (1,0), α (1, 0), ..., R (15, 0), G (15, 0), B (15, 0), α (15, 0), R (0, 1), G (0, 1) , B (0, 1), α (0, 1), R (1, 1), G (1, 1), B (1, 1), α (1, 1), ..., R (15, 1), G (15, 1), B (15, 1), α (15, 1), ..., R (0, 15), G (0, 15), B (0, 15), α ( 0, 15), R (1, 15), G (1, 15), B (1, 15), α (1, 15), ..., R (15, 15), G (15, 15), B (15, 15), α (15, 15)

  The second encoding unit 306 compresses the reversible compression target image into compressed data having a format shown in FIG. 9 as will be described later in units of 16 × 16 pixel rectangles. FIG. 4 shows an exemplary configuration of the second encoding unit 306.

  The boundary extraction unit 401 identifies a boundary where the color value changes and a boundary where the attribute value changes. This division boundary may be a boundary where at least one of the color value and the attribute value changes. In the present embodiment, the boundary extraction unit 401 divides an image area and generates boundary information indicating the position of a division boundary that divides each area. In the present embodiment, the boundary extraction unit 401 identifies a boundary where the color value changes and a boundary where the attribute value changes in the lossless compression region. One divided region in the lossless compression region defined by the boundary thus specified is composed of adjacent pixel groups having the same color value and attribute value. In addition, the boundary extraction unit 401 identifies a boundary where the attribute value changes in the lossy compression region. One divided area in the lossy compression area defined by the boundary thus specified is composed of adjacent pixel groups having the same attribute value. In this way, in the lossy compression region, the region delimited by the boundary where the attribute value changes may include pixels having different color values. The boundary extraction unit 401 divides the image area into one or more divided areas according to such rules.

  First, a specific processing example for the lossless compression area will be described below. The boundary extraction unit 401 extracts, as boundary pixels, pixels having different color values or attribute values from the upper left pixel for each line from the upper end. The pixel at the left end of the lossless compression area is always extracted as a boundary. Further, the boundary extraction unit 401 also extracts pixels having different color values or attribute values from the upper adjacent pixels as boundary pixels. Then, the boundary extraction unit 401 detects pixels having the same color value and attribute value in the processed upper adjacent line, and determines the boundary pixel in the current line and the boundary pixel in the upper adjacent line. Identify the boundaries by associating them. As described above, the boundary extraction unit 401 identifies a boundary where the color value does not change but the attribute value changes in addition to the boundary where the color value and the attribute value change.

  In the lossy compression region, the boundary extraction unit 401 extracts, as boundary pixels, pixels having different attribute values from the upper left pixel for each line from the upper end. Furthermore, the boundary extraction unit 401 also extracts pixels having different attribute values from the upper adjacent pixels as boundary pixels. The boundary extraction unit 401 detects pixels having the same attribute value in the processed upper adjacent line, and associates the boundary pixel in the current line with the boundary pixel in the upper adjacent line. Identify the boundaries.

  The header generation unit 402 generates information indicating the color value and attribute value of the region delimited by the boundary specified by the boundary extraction unit 401. However, for the lossy compression area, information indicating the color of the area is not generated, and information indicating the attribute value of the area delimited by the boundary specified by the boundary extraction unit 401 is generated. In addition, the trajectory generation unit 403 generates boundary position information specified by the boundary extraction unit 401. If the information generated by the header generation unit 402 and the trajectory generation unit 403 includes information indicating the color value and attribute value of the region delimited by the boundary or the boundary position information specified by the boundary extraction unit 401, It is not limited to a specific format. For example, information indicating the color value and attribute value of an area existing on the right side of the boundary and boundary position information may be generated. Furthermore, information indicating the color value and attribute value of an area sandwiched between a set of boundaries and position information of each boundary may be generated.

  In the present embodiment, information indicating the start boundary that is the left end boundary and the end boundary that is the right end boundary is generated for each of the lossless compression regions partitioned by the boundary specified by the boundary extraction unit 401. As described above, the boundary extraction unit 401 specifies an area that is defined by the set of the start boundary and the end boundary and has a constant color value and attribute value between the start boundary and the end boundary.

  Then, the header generation unit 402 generates information indicating the color value and the attribute value of the pixels in the area for each area defined by the combination of the start boundary and the end boundary. Further, the header generation unit 402 further generates start point coordinate information and end point coordinate information of the start boundary and the end boundary. As a specific example, the header generation unit 402 generates information indicating the X and Y coordinates at the uppermost end of the start boundary, the Y coordinate at the lowermost end of the start boundary, and the X coordinate at the uppermost end of the end boundary. On the other hand, the trajectory generation unit 403 generates trajectory information by encoding and aligning the coordinate values of the coordinates through which the start boundary and the end boundary pass for each of the start boundary and the end boundary. For example, the trajectory generation unit 403 generates information indicating the displacement of the X coordinate of the boundary between each horizontal line from the uppermost end to the lowermost end as trajectory information.

  Further, in the present embodiment, for each lossy compression region delimited by the boundary specified by the boundary extraction unit 401, information indicating the boundary where the attribute value changes is generated. As a specific example, for each irreversible compression region delimited by a boundary where the attribute value specified by the boundary extraction unit 401 changes, information indicating the leftmost boundary (attribute boundary) is generated. Then, the header generation unit 402 generates information indicating the changed attribute value at the boundary where the attribute value changes. As a specific example, the header generation unit 402 generates information indicating the attribute value of the pixel in the region for the region on the right side of the attribute boundary. The header generation unit 402 further generates starting point coordinate information and end point coordinate information of the attribute boundary. As a specific example, the header generation unit 402 generates information indicating the X coordinate and the Y coordinate at the uppermost end of the attribute boundary and the Y coordinate at the lowermost end of the attribute boundary. Then, the trajectory generation unit 403 generates trajectory information by encoding the coordinate values of the coordinates through which the attribute boundary passes.

  On the other hand, when the attribute value of the irreversible compression area divided by the boundary specified by the boundary extraction unit 401 is the same as the attribute value of the left-side lossless compression area, information indicating the attribute boundary for this area is omitted. Is done. In the present embodiment, the information indicating the attribute boundary is omitted using the information indicating the end boundary of the left-side lossless compression area and the start boundary of the right-side lossless compression area or the attribute boundary of the lossy compression area. The irreversible compression area can be specified.

  Information generated by the header generation unit 402 as described above is stored in the first area (header portion) of the lossless compressed data 901. The information generated by the trajectory generation unit 403 is stored in the second area (trajectory data unit) of the lossless compressed data 901.

  Next, the format of the compressed data generated by the system shown in FIG. 1 will be described with reference to FIGS.

  FIG. 5 is a memory image of compressed bitmap data included in the intermediate data. The memory image 5 includes a bitmap header 501 and compressed bitmap data 502 as bitmap data included in a drawing area such as a page or a frame.

  The bitmap header 501 is a set of headers defined for each bitmap (bitmap 1 to bitmap B in the example of FIG. 5) included in a drawing area such as a page or a frame. The bitmap header includes a bitmap size 5011, the number of channels 5012, a bit precision 5013, a pointer map 5014 to lossless compressed data, and a pointer map 5015 to lossy compressed data.

  One bitmap is divided into rectangular regions of 16 × 16 pixels. In the following description, it is assumed that the bitmap 1 is composed of W × H rectangular areas. If the number of vertical or horizontal pixels in the original bitmap is not an integer multiple of 16, padding is added so that the number of vertical and horizontal pixels is an integer multiple of 16, and then the bitmap is divided into rectangular areas can do.

  The bitmap size 5011 indicates the size of the bitmap, and more specifically indicates how many times it is 16 for each of the width and height of the bitmap. The channel number 5012 indicates the number of colors in the bitmap. The number of channels 5012 is 3 channels for RGB, 4 channels for CMYK, 3 channels for YUV, and 1 channel for BW. The bit precision 5013 indicates how many bits are expressed per channel. In the pointer map 5014 to the lossless compressed data, the head address of the lossless compressed data corresponding to each rectangular area is stored. The pointer map 5015 to the lossy compressed data stores the head address of the lossy compressed data corresponding to each rectangular area.

  The compressed bitmap data 502 is obtained by compressing a bitmap (bitmap 1 to bitmap B) included in a drawing area such as a page or a frame. The lossless compressed data is composed of lossless compressed data for each bitmap, and the lossy compressed data is also composed of lossy compressed data for each bitmap. Further, the lossless compressed data for one bitmap is composed of lossless compressed data corresponding to each rectangular area, and the irreversible compressed data for one bitmap is also composed of irreversible compressed data corresponding to each rectangular area. Is done.

  For example, the compressed bitmap data of bitmap 1 includes lossless compression data 5016, 5017,..., 5018 for each rectangular area and lossy compression data 5019, 5020,. The compressed bitmap data of the rectangular area 1 in the bitmap 1 includes lossless compressed data 5016 and lossy compressed data 5019. The lossless compressed data of each rectangular area includes boundary information in the rectangular area of 16 × 16 pixels, and color information and attribute information about each divided area (monochromatic area). The irreversible compressed data of each rectangular area is 16 × 16 pixel JPEG data composed of a total of four 8 × 8 pixel MCUs (Minimum Coded Units), two vertically and horizontally. The lossless compressed data 5016, 5017,..., 5018 and the lossy compressed data 5019, 5020,..., 5021 will be further described with reference to FIGS.

  FIG. 6 shows a rectangular area 6 of 16 × 16 pixels included in one bitmap. In FIG. 6, the solid line indicates the outer edge of the rectangular area and the boundary between the divided areas, and the broken line indicates the boundary of the 8 × 8 small rectangular area corresponding to the JPEG MCU. FIG. 6 also shows divided regions 601 to 608 that are separated by boundaries. Specifically, FIG. 6 shows a first area 601 (lossless compression area), a second area 602 (lossy compression area), a third area 603 (lossless compression area), and a fourth area 604. (Reversible compression area) is shown. FIG. 6 further shows a fifth area 605 (lossy compression area), a sixth area 606 (lossless compression area), a seventh area (lossless compression area), and an eighth area 608 (lossy compression). Area) is shown.

  In the rectangular area 6, in the second area 602, the fifth area 605, and the eighth area 608, three or more pixels having the same color value and attribute value are not adjacent to each other. Lossy compression. In the first region 601, the third region 603, the fourth region 604, the sixth region 606, and the seventh region 607, three or more pixels having the same color value and attribute value are adjacent to each other. This area is reversibly compressed.

  FIG. 7 shows a boundary (start boundary) having an attribute indicating the head on the horizontal line of the lossless compression area and a boundary having an attribute indicating the end on the horizontal line of the lossless compression area (dividing the rectangular area 6). Terminal boundary). In FIG. 7, the solid line indicates the start boundary, and the thick broken line indicates the end boundary. One lossless compression area starts from the pixel right next to the start boundary, and one lossless compression area ends at the pixel right next to the end boundary. Hereinafter, a set of a start boundary and an end boundary representing one lossless compression area is referred to as a lossless compression area boundary. The start boundary and the end boundary are generated by the second encoding unit 306 as boundary information that forms a pair so as not to cross each other.

  FIG. 7 includes a start boundary 701 and a termination boundary 702 of the first region 601, a start boundary 703 and a termination boundary 704 of the third region 603, a start boundary 705 and a termination boundary 706 of the fourth region 604, It is shown. FIG. 7 further shows the start boundary 707 and the end boundary 708 of the sixth area 606 and the start boundary 709 and the end boundary 710 of the seventh area 607.

  FIG. 8 shows a boundary that defines only attributes for the lossy compression region in the rectangular region 6. FIG. 8 shows a boundary 801 that defines the attribute of the second area 602, a boundary 802 that defines the attribute of the fifth area 605, and boundaries 803 and 804 that define the attribute of the eighth area 608. In the area 602 and the fifth area 605 in FIG. 2, the attribute does not change in the area. On the other hand, the eighth area 608 includes two areas having different attributes. Hereinafter, a boundary that defines only these attributes is referred to as an attribute boundary.

  FIG. 9 shows a format of lossless compressed data. Reference numeral 901 denotes lossless compression data of a rectangular area of 16 × 16 pixels, which corresponds to one of the lossless compression data 5016, 5017,. The lossless compressed data 901 includes a header portion (first area) in which the boundary number 9011 and boundary headers 9012,..., 9013 are stored, and a locus data portion (second area) in which the locus data 9014,. The area is divided into two areas).

  The header portion of the lossless compressed data 901 includes a color value and an attribute value indicating a region delimited by a boundary where the color value changes and a boundary where the attribute value changes in an image including pixels having color values and attribute values. 1 data element is included. For example, information generated by the header generation unit 402 can be stored in the header part. The trajectory data portion of the lossless compressed data 901 includes a second data element indicating the position of the boundary. For example, information generated by the trajectory generation unit 403 can be stored in the trajectory data unit.

  The number of boundaries 9011 indicates the total value of the number of sets of the start boundary and the end boundary of the lossless compression region and the number of attribute boundaries that define only the attributes in the 16 × 16 pixel rectangular region. In FIG. 9, the number of boundaries is E. That is, there are first to Eth boundaries in the rectangular area, and each boundary is a lossless compression area boundary or an attribute boundary. Hereinafter, the Nth boundary is referred to as boundary N-1.

  The header portion of the lossless compressed data 901 stores a boundary header that is header information of the boundary 0 to the boundary E-1. The header part is divided into a plurality of areas, and each divided area is defined by using the color value and attribute value of one reversible compression area defined by the combination of the start boundary and the end boundary, or the attribute boundary. The attribute value of one lossy compression area to be stored is stored.

  For example, a boundary header 9012 at boundary 0 is a boundary header at the first boundary, and a boundary header 9013 at boundary E-1 is a boundary header at the Eth boundary. In the trajectory data portion of the reversible compression data 901, trajectory data of the boundary 0 to the boundary E-1 is stored. For example, the trajectory data 9014 is trajectory data at the boundary 0, and the trajectory data 9015 is trajectory data at the boundary E-1. This trajectory data is data for specifying the horizontal (X) coordinate of the boundary in each horizontal line. The function, the horizontal (X) coordinate on the horizontal line, and the horizontal direction (X ) Coordinate difference, etc.

  Reference numeral 902 denotes the format of the boundary header for the lossless compression region boundary, that is, the boundary header corresponding to the set of the start boundary and the end boundary of the lossless compression region. A region color / attribute value 9021 indicates a color value and an attribute value of the lossless compression region, respectively. The boundary type 9026 indicates whether the boundary is a lossless compression boundary or an attribute boundary, and indicates that the boundary is a lossless compression boundary in the boundary header 902 for the lossless compression region boundary. As described above, information indicating whether the boundary is a boundary in the lossless compression area or the boundary in the lossy compression area is stored in the header portion of the lossless compression data 901.

  In addition, a part of locus position information can be stored in the boundary header. For example, the head Y coordinate 9022 indicates the vertical (Y) coordinate of the start point of the boundary, that is, the position of the uppermost end of the start boundary. The start X coordinate 9023 indicates the horizontal (X) coordinate of the start point of the start boundary, that is, the X coordinate at the uppermost end of the start boundary. An end Y coordinate 9024 indicates the vertical (Y) coordinate of the end point of the boundary, that is, the position of the lowest end of the start boundary. The end X coordinate (Disabling X Position) 9025 indicates the horizontal (X) coordinate of the start point of the end boundary, that is, the X coordinate at the uppermost end of the end boundary. Note that the start point of the end boundary and the vertical direction (Y) coordinates of the end point are the same as the start point and end point of the paired start boundary and the vertical direction (Y) coordinates, and thus are omitted in the boundary header 902. On the other hand, these pieces of position information can also be recorded as trajectory data.

  Reference numeral 903 denotes the format of the boundary header for the attribute boundary. The attribute value 9031 indicates the attribute value for the area on the right side of the attribute boundary. A starting X coordinate 9033 indicates the horizontal (X) coordinate of the starting point of the attribute boundary. Further, the boundary header 903 includes a head Y coordinate 9022 and a tail Y coordinate 9024 of the attribute boundary.

  The boundary headers 9012 to 9013 are sorted in ascending order of the leading Y coordinate 9022, and the boundary headers having the same leading Y coordinate 9022 are sorted in the ascending order of the starting Y coordinate 9023. That is, the boundary headers are stored in the memory in a state where the rectangular areas are arranged in the order of scanning from the upper left to the lower right.

  Below, the decoding method of the lossless compression data 901 is demonstrated. In each horizontal line, pixels between the horizontal (X) coordinate of the start boundary and the horizontal (X) coordinate of the end boundary for one lossless compression region boundary are the region colors defined in the boundary header 902. The pixel has the color value and the attribute value. On one horizontal line, the absence of a gap between the end boundary for one lossless compression region boundary and the start boundary for another lossless compression region boundary indicates that the lossless compression regions are adjacent.

  Also, on one horizontal line, there is a gap between the end boundary for one lossless compression region boundary and the start boundary for another lossless compression region boundary. The pixels up to the left of the start boundary are included in the lossy compression region. The color values of the pixels included in such a lossy compression area can be obtained by decompressing the lossy compression data. Further, the attribute value of the pixel included in such a lossy compression area is the same value as the left-side lossless compression area if there is no attribute boundary. On the other hand, when an attribute boundary is adjacent to the right of the end boundary for one lossless compression region boundary on one horizontal line, or when an attribute boundary exists in the lossy compression region, the attribute boundary The attribute values of the pixel and the pixel on the right side thereof are attribute values defined in the boundary header 903.

  When the start boundary of the lossless compression region boundary does not exist at the left end (X = 0) of the horizontal line, the left end of the lossless compression region boundary that appears first from the left end (X = 0) on the horizontal line. These pixels are included in the lossy compression region. When the start coordinate of the lossless compression region boundary does not exist in one horizontal line, the irreversible compression region includes the left end (X = 0) to the right end (X = 15). In this case, the attribute value of the pixel included in the irreversible compression area is the same value as the attribute value of the rightmost pixel on the horizontal line one line up if there is no attribute boundary. On the other hand, when there is an attribute boundary at the left end (X = 0) of the horizontal line, the attribute values of the left end pixel and the right side pixel are attribute values defined in the boundary header 903 of the attribute boundary.

  Note that either the start boundary of the lossless compression area boundary or the attribute boundary exists at the left end of the top line of the rectangular area.

  Table 1 shows lossless compression data for the rectangular area 6.

  Table 1 will be briefly described. The rectangular area 6 includes five lossless compression area boundaries corresponding to the first area, the third area, the fourth area, the sixth area, and the seventh area, which are lossless compression areas, and four attributes. Since there is a boundary, the boundary number (E) is nine.

  In the boundary header of boundary 0 which is the boundary for the first area, information indicating that the boundary type is a lossless compression area boundary, 0xFF0000 which is the color value of the first area as the area color, and the first as the attribute value 0, which is the attribute value of the area, is recorded. In addition, since the uppermost ends of the start boundary 701 and the end boundary 702 exist on the horizontal line of coordinates (Y = 0), the leading Y coordinate is 0, and the lowermost end is on the horizontal line of coordinates (Y = 9). Since it exists, the end Y coordinate is 9. Further, in the horizontal line of the coordinate (Y = 0) where the uppermost ends of the start boundary 701 and the end boundary 702 exist, the horizontal (X) coordinate of the start boundary 701 is 0, so the start X coordinate is 0, Since the lateral (X) coordinate of the end boundary 702 is 2, the end X coordinate is 2.

  In the boundary header of boundary 1 representing the attribute boundary 801, information indicating that it is an attribute boundary is recorded as the boundary type, and 2 that is the attribute value of the pixel in the right region of the attribute boundary 801 is recorded as the attribute value. Further, since the uppermost end of the attribute boundary 801 exists in the horizontal line of coordinates (Y = 0), the top Y coordinate is 0, and the lowermost end exists in the horizontal line of coordinates (Y = 3). The end Y coordinate is 3. Furthermore, since the horizontal (X) coordinate of the attribute boundary 801 is 3 in the horizontal line of the coordinate (Y = 0) where the uppermost end of the attribute boundary 801 exists, the start X coordinate is 3. Since the same applies to the boundaries 2 to 8, the description thereof is omitted.

  The trajectory data of the start boundary and the trajectory data of the end boundary are recorded as the trajectory data of the boundary 0 that is the boundary for the first region. The horizontal (X) coordinates of the start boundary from the start Y coordinate (Y = 0) to the end Y coordinate (Y = 9) are 0, 0, 0, 0, 0, 0, 1, 2, 3, 4 in order. It is. In this embodiment, 0, 0, 0, 0, 0, +1, +1, +1, +1, which are differences in the horizontal (X) coordinates between adjacent horizontal lines, are recorded as the trajectory data of the start boundary. . Similarly, the lateral direction (X) coordinates of the end boundary from the start Y coordinate (Y = 0) to the end Y coordinate (Y = 9) are 2, 5, 5, 6, 7, 6, 6, 6, 5 and 5. Therefore, +3, 0, +1, +1, −1, 0, 0, −1, 0, which is the difference in the horizontal (X) coordinate between adjacent horizontal lines, is recorded as the trajectory data of the end boundary. .

  As the locus data of the boundary 1 representing the attribute boundary 801, the locus data of the attribute boundary 801 is recorded. The horizontal (X) coordinates of the start boundary from the start Y coordinate (Y = 0) to the end Y coordinate (Y = 3) of the attribute boundary 801 are 3, 6, 6, and 7, respectively. Therefore, as trajectory data of the boundary 1 representing the attribute boundary 801, +3, 0, and +1 that are differences in the horizontal (X) coordinates are recorded. Since the same applies to the boundaries 2 to 8, the description thereof is omitted.

  Next, lossy compressed data will be described. The lossy compressed data in the rectangular area 6 is data compressed by JPEG, and is composed of four MCUs each having 8 × 8 pixels, and two MCUs are arranged vertically and horizontally. The irreversible compressed data indicates the color values of all the pixels in the rectangular area 6. Therefore, the first region 601, the third region 603, the fourth region 604, the sixth region 606, and the seventh region 607, which are lossless compression regions, also have color values.

  FIG. 10 shows an exemplary configuration of the second decoding unit 111 and the first output unit 112. In FIG. 10, the second decoding unit 111 includes a data reading unit 1001, a boundary tracking unit 1003, a span information generation unit 1004, a region color / attribute value reading unit 1005, an attribute value reading unit 1006, and an attribute value combining unit 1007. Prepare. The first output unit 112 includes an image reading unit 1002 and a pixel selection unit 1008.

  In the following, a case where decoding is performed on the rectangular area 6 of the bitmap 1 will be described. By sequentially decoding W × H rectangular areas from the rectangular area 1 to the rectangular area W × H, The whole can be decoded. Further, by performing such processing for each of the bitmaps 1 to B, the entire image can be decoded.

  The data reading unit 1001 reads lossless compressed data of the rectangular area 6 of the bitmap 1 included in the compressed bitmap data 502 from the RAM 102 in accordance with an instruction from the control unit 109. The image reading unit 1002 reads image data obtained by the first decoding unit 110 decoding the lossy compressed data of the rectangular area 6 of the bitmap 1 included in the compressed bitmap data 502.

  The boundary tracking unit 1003 and the span information generation unit 1004 generate span information with reference to the trajectory data stored in the trajectory data portion of the lossless compressed data 901. The span information indicates the position of the pixel group having the same color value and attribute value on the horizontal line of the image for the lossless compression region. The span information indicates the position of the pixel group having the same attribute value on the horizontal line of the image for the lossy compression region. As described above, in the locus data, position information on the boundary where the color value changes and the boundary where the attribute value changes is recorded.

  In the case of the present embodiment, the boundary tracking unit 1003 further acquires the start boundary and the end boundary of the lossless compression region boundary and the start point and end point of the attribute boundary for each of the boundary 0 to the boundary E-1 by referring to the boundary header. To do. Then, the boundary tracking unit 1003 determines the positions of the start boundary and the end boundary of the lossless compression region boundary and the position of the attribute boundary for each of the boundary 0 to the boundary E-1 by updating the coordinate value with reference to the trajectory data. And hold.

  The span information generation unit 1004 calculates the number of pixels in the region on the horizontal line with reference to the X coordinate of the boundary existing on each horizontal line. The span information generation unit 1004 generates span information including the X coordinate of the left end boundary and the number of pixels for each of the regions on the horizontal line. Based on the boundary header, the span information generation unit 1004 can know whether each piece of span information is span information of a lossless compression area or irreversible compression area.

  The area color / attribute value reading unit 1005 refers to the boundary header stored in the header part of the lossless compressed data 901 and determines the color value and attribute value of the pixel group indicated in the span information. As described above, information indicating the color value and attribute value of the area delimited by the boundary is recorded in the boundary header. Specifically, the area color / attribute value reading unit 1005 reads the color value and attribute value corresponding to the span information from the boundary header for each lossless compression area.

  The attribute value reading unit 1006 reads the attribute value corresponding to the span information from the boundary header for each lossy compression area. The attribute value combining unit 1007 combines the color value in the image read by the image reading unit 1002 and the attribute value read by the attribute value combining unit 1007, so that the color value and the attribute value are added to the pixel corresponding to the span information. Set.

  The pixel selection unit 1008 selects a pixel in the lossless compression region and a pixel in the lossy compression region based on the span information, and outputs the color value and attribute value of the sequentially selected pixels. Thus, the decoding of the lossless compressed data and the lossy compressed data is completed.

  The image processing method in the present embodiment will be briefly described with reference to the flowchart of FIG. FIG. 11A is a flowchart of the encoding process. In step S1110, the region dividing unit 302 divides an input image composed of pixels having color values and attribute values into a lossless compression region and a lossy compression region as described above. In step S1115, the boundary extraction unit 401 identifies a boundary where the color value changes and a boundary where the attribute value changes in the input image. In step S1120, the header generation unit 402 stores information indicating the color value and attribute value of the region delimited by the boundary in the header portion of the lossless compressed data. In step S1125, the trajectory generation unit 403 stores the boundary position information in the trajectory data portion of the lossless compressed data. In step S1130, the first encoding unit 304 generates lossy compressed data representing an input image in the lossy compression region.

  FIG. 11B is a flowchart of the decoding process. In step S1150, the data reading unit 1001 acquires lossless compressed data. In step S1155, the span information generation unit 1004 refers to the position information of the boundary where the color value changes and the boundary where the attribute value changes stored in the trajectory data part of the lossless compression data. Then, the span information generation unit 1004 generates span information indicating the position of the pixel group having the same color value and attribute value on the horizontal line of the image for the lossless compression region. Further, the span information generation unit 1004 generates span information indicating the position of a pixel group having the same attribute value on the horizontal line of the image for the lossy compression region. In step S1160, the region color / attribute value reading unit 1005 refers to the information indicating the color value and attribute value of the region delimited by the boundary stored in the header portion of the lossless compressed data. Then, the region color / attribute value reading unit 1005 determines the color value and attribute value of the pixel group indicated in the span information for the lossless compression region. In step S1165, the attribute value reading unit 1006 determines the attribute value of the pixel group indicated in the span information for the lossy compression area. In step S1170, the first decoding unit 110 generates image data by decoding the lossy compressed data. In step S1175, the attribute value combining unit 1007 corresponds to the span information by combining the color value in the image data obtained in step S1170 and the attribute value determined in step S1165 for the lossy compression region. A color value and an attribute value are set for the pixel.

  As described above, the image data is divided into the lossless compression area and the lossy compression area according to the configuration shown in FIGS. 3 and 4, and compressed data is generated. Also, with the configuration of FIG. 10, the respective compressed data is decoded and then combined to decode the image.

  Since the pixel selection unit 1008 outputs the color value and the attribute value together for each pixel, when performing image processing using the decoded image, the cache can be effectively used to perform processing at high speed. it can. Even during the decoding process, since both the color value and the attribute value of the lossless compression area are stored in the boundary header, the color value and the attribute value are held in the same memory area, and the area color / attribute value reading unit The processing by 1005 can be performed at high speed by effectively using the cache. As described above, the use of the reversible compression data 901 shown in FIG. 9 can improve the image processing performance.

  Further, according to the method of the present embodiment, information indicating the color value and attribute value of the region delimited by the boundary is stored in the header portion of the lossless compressed data. For this reason, the color value and attribute value of the image in the lossless compression area can be easily corrected. Also, the attribute value of the image in the lossy compression area can be easily corrected. According to such a configuration, for example, when an internal parameter of the printer changes, the color value and attribute value of the image included in the intermediate data can be changed with a small processing load in accordance with the change.

  110: 1st decoding part, 111: 2nd decoding part, 112: 1st output part, 304: 1st encoding part, 306: 2nd encoding part, 401: Boundary extraction part, 402: Header generation unit, 403: locus generation unit

Claims (15)

A specifying means for specifying a boundary where a color value changes and a boundary where an attribute value changes in an input image including pixels having a color value and an attribute value;
First compression for generating and outputting data including a first area for storing information indicating color values and attribute values of an area delimited by the boundary, and a second area for storing position information of the boundary Means,
An image processing apparatus comprising:   The image processing apparatus according to claim 1, wherein the specifying unit specifies a boundary where the color value does not change but the attribute value changes in addition to the boundary where the color value and the attribute value change. The specifying means specifies a region defined by a set of a start boundary and an end boundary, and having a constant color value and attribute value between the start boundary and the end boundary,
The first region has a plurality of divided regions;
The first compression means stores a color value and an attribute value of an area defined by a set of the start boundary and end boundary in one divided area of the first area, and the start boundary and end terminal The image processing apparatus according to claim 1, wherein boundary position information is stored in the second area. The first compression means stores start point coordinate information and end point coordinate information of the start boundary and end boundary in the first region, and stores trajectory information of the start boundary and end boundary in the second region. The image processing apparatus according to claim 3, wherein the image processing apparatus is stored.   The image processing apparatus according to claim 4, wherein the trajectory information of the start boundary and the end boundary is obtained by encoding coordinate values of coordinates through which the start boundary and the end boundary pass. Setting means for setting a reversible compression region and a lossy compression region in the input image;
Second compression means for generating and outputting lossy compressed data representing the input image in the lossy compression region;
The specifying means specifies a boundary where a color value changes and a boundary where an attribute value changes in the lossless compression region,
The first compression means stores information indicating the color value and attribute value of the lossless compression area in the first area, and outputs the data as lossless compression data. The image processing apparatus according to any one of 1 to 5. The specifying means specifies a boundary where the attribute value changes in the lossy compression region,
The first compression unit further stores, in the first area, information indicating an attribute value of an area delimited by the boundary for the lossy compression area, and stores the boundary position information in the second area. The image processing apparatus according to claim 6, wherein the image processing apparatus is stored.   The first compression means stores the attribute value after the change at the boundary where the attribute value changes, the starting point coordinate information and the end point coordinate information of the boundary where the attribute value changes, in the first area, and The image processing apparatus according to claim 7, wherein trajectory information of a boundary where the attribute value changes is stored in the second area.   The first area further stores information indicating whether the boundary is a boundary in the lossless compression area or a boundary in the lossy compression area. The image processing apparatus described.   The second compression unit generates the lossy compressed data by irreversibly compressing the input image after replacing a color value of a pixel in the lossless compression region with a predetermined value. The image processing apparatus according to any one of claims 7 to 9. In an image composed of pixels having color values and attribute values, a first data element indicating a color value and an attribute value of a region delimited by a boundary where the color value changes and a boundary where the attribute value changes is a first region. Including
Data having a structure including a second data element indicating the position of the boundary in a second area. An acquisition unit that acquires data output from the image processing apparatus according to any one of claims 1 to 10, or data according to claim 11.
Pixels having the same color value and attribute value on the horizontal line of the image with reference to the position information of the boundary where the color value changes and the boundary where the attribute value changes stored in the second area of the data Information generating means for generating span information indicating the position of the group;
A decision to determine a color value and an attribute value of a pixel group indicated in the span information with reference to information indicating a color value and an attribute value of an area delimited by the boundary stored in the first area of the data; Means,
An image processing apparatus comprising: An image processing method performed by an image processing apparatus,
A specifying step of identifying a boundary where a color value changes and a boundary where an attribute value changes in an input image composed of pixels having a color value and an attribute value;
A compression step of generating and outputting data including a first area storing information indicating color values and attribute values of an area delimited by the boundary, and a second area storing position information of the boundary;
An image processing method comprising: An image processing method performed by an image processing apparatus,
An acquisition step of acquiring the data according to claim 11 or the data output by the image processing method according to claim 13;
Pixels having the same color value and attribute value on the horizontal line of the image with reference to the position information of the boundary where the color value changes and the boundary where the attribute value changes stored in the second area of the data An information generating step for generating span information indicating the position of the group;
A decision to determine a color value and an attribute value of a pixel group indicated in the span information with reference to information indicating a color value and an attribute value of an area delimited by the boundary stored in the first area of the data; Process,
An image processing method comprising:   A program for causing a computer to function as each unit of the image processing apparatus according to any one of claims 1 to 10 and claim 12.

Images