JP4748805B2 - Image coding apparatus and control method thereof - Google Patents

Description

  The present invention relates to a technique for compressing and encoding image data.

  Conventionally, as a method for efficiently encoding an image in which a character or a line drawing (hereinafter referred to as a character line drawing) and a natural image are mixed, a proposal has been made in Patent Document 1. This technique is briefly described as follows.

First, multi-value image data to be encoded is divided into blocks that are units of orthogonal transformation. Since the character line drawing has the same color, the most frequent pixel in the block is used as the character line drawing, and the most frequent color information is extracted (hereinafter referred to as an extracted color). Then, binary identification information for distinguishing the pixel determined to be the extracted color in the target block from the pixel determined to be the non-extracted color is generated. Then, an average value of pixels determined as non-extracted pixels is obtained, and the pixel value determined as the extracted color is replaced with the average value. That is, all the pixels in the block of interest are pixels that constitute a natural image. The information indicating the extracted color and the identification information are losslessly encoded. On the other hand, the irreversible JPEG encoding for performing orthogonal transformation is performed on the target block after the replacement processing. Thereafter, these three encoded data are multiplexed to generate encoded data for one block. As a result, the high-frequency component and edge component of the character / line image are reduced in the JPEG encoding processing target block, and the encoding efficiency is increased accordingly.
JP-A-4-326669

  However, according to the above-mentioned patent document 1, the extracted color is determined independently for each block, so that the extracted color may be different between adjacent blocks. In such a situation, as a result, the quality differs between adjacent blocks in the decoded image, and an unnatural image is reproduced.

  Therefore, the present invention intends to provide a technique for increasing the probability that the extracted colors in the adjacent blocks are the same color, so that the adjacent blocks in the decoded image are naturally connected to each other.

In order to solve this problem, for example, an image encoding device of the present invention has the following configuration. That is,
An image encoding device for encoding multi-value image data,
Input means for inputting image data in block units to be subjected to lossy encoding from multi-valued image data to be encoded;
Extraction means for extracting color information indicating the color of the pixel of the line drawing in the image data of the input block;
Position information generating means for generating position information for identifying the position of the pixel of the line drawing in the target block and the position of the pixel of the non-line drawing according to the color information extracted by the extracting means;
Lossless encoding means for losslessly encoding the position information generated by the position information generation means and the color information extracted by the extraction means;
In order to replace the value of a line drawing pixel according to the position information generated by the position information generation means, a replacement value calculation means for calculating an average value of the non-line drawing pixel value as a replacement value;
Replacement means for replacing the value of the pixel of the line drawing in the block of interest with the replacement value;
Irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplex means for multiplexing and outputting encoded data of position information and color information generated by the lossless encoding means, and encoded data of image data generated by the lossy encoding means,
The extraction means includes
Storage means for storing color information extracted from already encoded blocks;
Candidate determination means for determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination unit that refers to the storage unit and determines whether there is a color that matches color information of an adjacent block among the extracted color candidates determined by the candidate determination unit;
Selection means for preferentially selecting color information determined by the determination means in the extracted color candidates as color information of a line drawing in the block of interest ;
When the determination unit determines that there is no color that matches the color information of an adjacent block among the extraction color candidates determined by the candidate determination unit, the selection unit selects the extraction color candidate having the maximum frequency as the target block. Select as the color information of the line drawing in
When the determining means determines that there is a color that matches the color information of the adjacent block among the extracted color candidates determined by the candidate determining means, the selecting means The color information that maximizes the number of appearances by the combination of the color information extracted in the adjacent block and the extracted color candidate in the target block is selected as the color information of the line drawing in the target block .

  According to the present invention, a color that matches the color information determined in the adjacent block is preferentially selected from the color candidates determined as the line drawing in the block of interest. As a result, it is possible to generate encoded data for reproducing an image in which the color variation of the line drawing of the adjacent block is suppressed. In addition, when performing lossy encoding of the gradation image of the block of interest, encoded data by lossy encoding is performed by replacing the color of a pixel determined to be a line drawing with the color of a pixel determined to be a non-line drawing. The size can also be reduced.

  Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

<First Embodiment>
FIG. 1 shows a block configuration diagram of an image encoding device according to the first embodiment. This apparatus includes a blocking unit 100, an extracted color determination unit 110, a buffer 120, a delay unit 130, a comparison unit 140, a replacement value generation unit 150, a selector 160, three encoding units 170 to 172, and a multiplexing unit 180. Composed.

  The block forming unit 100 has a memory for several lines inside, and 8 × 8 from multi-valued color image data (R, G, B each having 8 bits) input in raster order from an input terminal (not shown). A block composed of pixels is cut out. Then, the blocking unit 100 outputs the extracted block image data to the extraction color determination unit 110 and the delay unit 130 (hereinafter referred to as block data). The order of the blocks to be output is also the raster order with the block as a unit.

  Note that the encoding unit 172 encodes a gradation image (multi-valued image) according to JPEG encoding, which is lossy encoding. In JPEG encoding, DCT conversion, quantization, and entropy encoding are performed in units of 8 × 8 pixels, so the blocking unit 100 sets the size of one block to 8 × 8 pixels. However, the size of one block may be 8m × 8n (m and n are positive integers). In this case, since there are m × n 8 × 8 pixels in one block, it is possible to perform JPEG encoding. The image data input source input by the blocking unit 100 is an image scanner, but may be a storage medium that stores image data to be encoded as a file. That is, the type of input source does not matter.

  The extracted color determining unit 110 refers to the buffer 120, determines the extracted color of the character line drawing of the block of interest from 8 × 8 pixel data in the input block, and compares the extracted color information with the comparing unit 140. Are output to the encoding unit 170 and stored in the buffer 120. The buffer 120 has a capacity for storing at least one line + 1 extracted color information, and erases the oldest extracted color information when new extracted color information is stored. Therefore, the buffer 120 can be composed of a FIFO memory. The reason for providing the buffer 120 is that when the extraction color determination unit 110 determines the extraction color of the block of interest, the extraction color of the already encoded block adjacent to the block of interest is referred to. Although details will be described later, the extracted color determination unit 110 outputs a control signal to the comparison unit 140 and the multiplexing unit 180 as to whether or not the extracted color exists in the target block.

  The encoding unit 170 performs lossless encoding on the extracted color information of the block of interest output from the extracted color determining unit 110 and outputs the encoded data to the multiplexing unit 180.

  The delay unit 130 is interposed between the blocking unit 100 and the comparison unit 140, delays the input block by the time required for the extraction color determination unit 110 to determine the extraction color, and then compares the comparison unit 140. , Output to the replacement value generation unit 150 and the selector 160. That is, the delay unit 130 is provided to synchronize processing on the block of interest in the comparison unit 140, the replacement value generation unit 150, and the selector 160.

  When the extracted color determining unit 110 receives a control signal indicating that there is an extracted color, the comparing unit 140 compares the extracted color with each pixel value in the block of interest, and a pixel position having the same color as the extracted color The position information for specifying is generated. Then, the comparison unit 140 outputs the generated position information to the replacement value generation unit 150, the selector 160, and the encoding unit 171. This position information is also information indicating whether or not it is the same as the extracted color, and is represented by binary 64-bit (= 8 × 8) data. In the embodiment, a bit of “1” is assigned to a pixel having the same color as the extracted color, and “0” is assigned to a pixel having a color different from the extracted color (hereinafter referred to as a non-extracted color).

  When receiving a control signal indicating no extracted color from the extracted color determining unit 110, the comparing unit 140 outputs position information in which all bits are “0”.

  The encoding unit 171 performs lossless encoding on the position information and outputs the encoded data to the multiplexing unit 180. In the embodiment, the encoding unit 171 performs run-length encoding, but the type thereof is not limited as long as it performs lossless encoding.

  The replacement value generation unit 150 calculates a replacement value for replacing the value of the pixel having the extracted color based on the 8 × 8 pixel image data via the delay unit 130 and the position information, and selects the replacement value as a selector. To 160. Specifically, the replacement value generation unit 150 calculates an average value of each color component of a pixel having a non-extracted color, and calculates an average value of each component value as a replacement value.

  The selector 160 inputs position information, replacement values, and block image data. The selector 160 outputs the pixel data at the corresponding position in the block as it is at the pixel position where the position information is “0”, and replaces the pixel value in the block at the pixel position where the position information is “1”. Output the replacement value. Therefore, when the extracted color determining unit 110 outputs a control signal indicating that there is no extracted color, all the bits of the position information are “0”. Therefore, the selector 160 directly inputs the input block data to the encoding unit 172. Will be output.

  Note that the control signal of the extracted color determination unit 110 may be output to the selector 160 instead of being output to the comparison unit 140. In this case, when the selector 160 receives a control signal indicating no extracted color, the processing equivalent to the above can be realized by selecting block data unconditionally.

  The encoding unit 172 performs JPEG encoding, which is lossy encoding, on the 8 × 8 pixel image data input via the selector 160, and outputs the resulting encoded data to the multiplexing unit 180.

  The multiplexing unit 180 combines the encoded data from the encoding units 170, 171, and 172 based on the control signal from the extracted color determination unit 110 to generate and output an encoded data sequence for one block. (Details will be described later).

  The configuration and processing of the encoding apparatus according to the embodiment are generally as described above. Hereinafter, the processing content of the characteristic configuration will be described in more detail.

  The processing of the extracted color determination unit 110 in the embodiment will be described with reference to FIGS.

  The extracted color determination unit 110 generates a histogram of colors in the input block data of interest and calculates the frequency of colors included in the block data. Then, a color having a number equal to or greater than a preset threshold value is determined as an extraction color candidate for the block of interest. If the number of colors equal to or greater than the threshold cannot be detected, a control signal indicating no extracted color is output to the comparison unit 140 and the multiplexing unit 180. If there is even one extracted color candidate, the extracted color determination unit 110 outputs a control signal indicating the presence of the extracted color to the comparison unit 140 and the multiplexing unit 180.

The extracted color determination unit 110 in this embodiment determines the extracted color of the block of interest as follows when the number of extracted color candidates is one or more.
(1) If none of the extracted color candidates of the target block matches the color extracted by the adjacent block, the extracted color determining unit 110 determines the most frequent color among the extracted color candidates as the extracted color of the target block. To do. FIG. 3A shows such a state. In the illustrated case, there are two extraction candidates for the target block {A, B}, and the frequency is A ≧ B. Further, it is shown that the extracted color of three adjacent blocks is either C or D. In such a situation, the color “A” is determined as the extracted color of the block of interest. When the frequencies of the colors A and B are the same, the one having the smaller color difference from the extracted color of the adjacent block is selected.
(2) If there is at least one extracted color candidate of the target block that matches the extracted color of the adjacent block, the extracted color determination unit 110 determines that the extracted color candidates match. As a result, the color that becomes the most frequent is determined as the extracted color of the target block.

  FIG. 3B shows an example in which there are two extraction candidates for the target block {A, B}, and the extracted colors of adjacent blocks are {B, C}. In this case, since only the color B is common, the extracted color determining unit 110 determines the color B as the extracted color of the block of interest.

  FIG. 3C shows that among the extraction candidates for the target block, there are two colors {A, B} that match the extracted colors of adjacent blocks. In this case, if the extraction color candidate “A” of the target block is temporarily selected, the number of blocks having the extraction color “A” is two. If the extraction color candidate “B” of the block of interest is selected, the number of blocks having the extraction color “B” is three. Therefore, in this case, the extracted color determining unit 110 determines the color “B” as the extracted color of the block of interest.

  A specific example of the processing content of the extracted color determination unit 110 in the embodiment will be described with reference to the flowchart of FIG. This figure shows a processing procedure for one block data in the extracted color determination unit 110.

  First, in step S <b> 21, the extracted color determination unit 110 obtains a histogram of pixel colors in the target block data. When creating this histogram, R, G, and B data are counted as one set.

  Next, in step S22, the extracted color determining unit 110 determines whether there is a color having a frequency equal to or higher than a preset threshold Th from the histogram. When it is determined that there is no color having a frequency exceeding the threshold, that is, when it is determined that there is no character line drawing pixel in the target block, the process proceeds to step S23, and a control signal indicating no extracted color is multiplexed. The data is output to 180, and this processing is finished.

  If it is determined in step S22 that there is a color having a frequency equal to or higher than the threshold, that is, if it is determined that a character / line drawing pixel exists in the block of interest, the process proceeds to step S24. In step S <b> 24, the extracted color determination unit 110 outputs a signal indicating the presence of the extracted color to the multiplexing unit 180. Hereinafter, a color having a frequency equal to or higher than the threshold (a plurality of colors) may be referred to as an extraction color candidate.

  In the next step S25, the extracted color determination unit 110 determines whether or not the extracted color candidate has the same color as the extracted color of the adjacent block immediately before, directly above, upper left, and upper right of the block of interest. This is determined by referring to the data stored in 120.

  If it is determined that the extracted color candidates of the target block are the same as the extracted colors of the adjacent blocks, the process proceeds to step S27. In this step S27, the most frequent color that matches the extracted color of the adjacent block is obtained from the extracted color candidates of the target block, and that color is determined as the final extracted color of the target block (FIG. 3B). ), (C)).

  If it is determined that there is no extracted color candidate of the target block that is the same as the extracted color of the adjacent block, the most frequent color among the extracted color candidates of the target block is selected as the target color in step S26. The final extracted color of the block is determined (see FIG. 3A).

  When the process of either step S26 or S27 is completed, the process proceeds to step S28. Here, the extracted color determination unit 110 outputs the RGB component data of the determined extracted color to the buffer 120, the comparison unit 140, and the encoding unit 170.

  As a result, in the input block data, the comparison unit 140 displays position information “1” at the pixel position that is the extraction color output from the extraction color determination unit 110 and “0” at the pixel position that is the non-extraction color. Can be generated.

  Next, the multiplexing unit 180 in the embodiment will be described.

The processing of the multiplexing unit 180 generates block-unit encoded data in accordance with the control signal from the extracted color determining unit 110. Specifically, there are the following two types.
(1) When a signal indicating the presence of an extracted color is received from the extracted color determining unit 110:
As shown in FIG. 4A, a header including a flag indicating that there is an extracted color (one bit is sufficient) is generated. Then, the header, the lossless encoded data of the extracted color information output from the encoding unit 170, the lossless encoded data of the position information output from the encoding unit 171 and the floor output from the encoding unit 172. The lossy encoded data of the key data is combined (multiplexed) to generate a code string. The multiplexing unit 180 stores the data size of each encoded data in the header.
(2) When a signal indicating no extracted color is received from the extracted color determining unit 110:
As shown in FIG. 4B, a header including a flag indicating no extracted color is generated. Then, the header and the lossy encoded data of the gradation data output from the encoding unit 172 are combined (multiplexed) to generate a code string. That is, the encoded data from the encoding units 170 and 171 are ignored. Note that the multiplexing unit 180 stores the data size of the encoded data of the gradation image in the header.

  The configuration and processing contents of the encoding device in the embodiment have been described above. Next, the decoding device will be described.

  FIG. 5 is a block diagram of the image decoding apparatus. This apparatus includes a buffer 50, a header analysis unit 51, a separation unit 52, decoding units 53 to 55, and a selector 56. The decoding unit 53 decodes the encoded extracted color information, and the decoding unit 54 decodes the encoded position information. The decoding units 53 and 54 correspond to the encoding units 170 and 171 in FIG. 1 and perform lossless decoding. The decoding unit 55 decodes the encoded gradation image, and corresponds to the encoding unit 172. In the embodiment, since the encoding unit 172 performs JPEG encoding, the decoding unit 55 performs a process of decoding JPEG encoded data.

  The buffer 50 stores the image data encoded by the encoding device. The encoded image data stored in the buffer 50 may be read from a storage medium or received via communication means. That is, the input source of encoded image data is not limited.

  The header analysis unit 51 analyzes the header of one block of encoded data stored in the buffer 50, determines whether the flag included in the header indicates that there is an extracted color or no extracted color, and the determination The result is output as a control signal to the separation unit 52 and the selector 56. In addition, the header analysis unit 51 outputs information indicating each encoded data size in the encoded data for one block to the separation unit 52 as a control signal.

  The separation unit 52 inputs encoded data excluding the header among the encoded data for one block stored in the buffer 50.

  In the case where the signal indicating the presence of the extracted color is input from the header analysis unit 51, the separation unit 52 also inputs the size information of each encoded data, so that each code is displayed at the position indicated by the size. Data is separated and distributed to the decoding units 53 to 55. As a result, the decoding unit 53 decodes the encoded extracted color information, the decoding unit 54 decodes the encoded position information, and the decoding unit 55 decodes the encoded gradation information. The decoding results in the decoding units 53 to 55 are output to the selector 56.

  Further, when the separation unit 52 receives a signal indicating no extracted color from the header analysis unit 51, the encoded data excluding the header is only the encoded data of the gradation image. Is output to the decoding unit 55. As a result, the decoding unit 55 decodes the encoded gradation information.

  When the control signal indicating the presence of the extracted color is input from the header analysis unit 51, the combining unit 56 selects and outputs the extracted color information from the decoding unit 53 at the pixel position where the position information is “1”. At the pixel position where the position information is “0”, the pixel value of the gradation image from the decoding unit 55 is selected and output.

  Further, when a control signal indicating no extracted color is input from the header analysis unit 51, the synthesis unit 56 unconditionally selects and outputs the pixel data output from the decoding unit 55.

  As described above, according to the present embodiment, the color of the pixel determined as the character / line drawing in the target block is determined with reference to the extracted color of the adjacent block. Therefore, the probability that the color of the pixels of the character / line drawing between the blocks will be the same, and it becomes possible to generate encoded data that becomes a natural character / line drawing image. In addition, since the encoding unit 172 performs irreversible JPEG reverse encoding on the assumption that there is no pixel in the character line image, the high-frequency component is inevitably reduced in the conversion coefficient obtained by the orthogonal conversion, High compression ratio can be expected.

<Second Embodiment>
In the first embodiment, when determining one of the extracted color candidates for the target block, the most frequent combination with the extracted color of the adjacent block is selected. However, it is more desirable that the colors of the character / line drawing pixel groups that are continuous across two adjacent blocks are the same. Therefore, in the second embodiment, an example for realizing this will be described.

  FIG. 6 shows the configuration of the encoding apparatus according to the second embodiment. The difference from FIG. 1 is that the processing content of the extracted color determination unit 110 and the position information from the comparison unit 140 are also stored in the buffer 120. Note that the buffer 120 stores the extracted color of each block that has already been encoded and its position information. Other configurations related to the encoding process and the configuration of the decoding apparatus are the same as those in the first embodiment. Therefore, hereinafter, the process of the extracted color determination unit 110 in the second embodiment will be described.

The extraction color determination unit 110 in the second embodiment determines the extraction color of the block of interest as follows.
(1) When there is no extraction candidate for the target block that is the same as the extraction color of the adjacent block, the most frequent color among the extraction color candidates for the target block is determined as the final extraction color of the target block. That is, it is the same as the first embodiment (see FIG. 3A).
(2) If there is only one extraction candidate for the target block that has the same extracted color as the adjacent block, that color is determined as the final extracted color of the target block. This is also the same as in the first embodiment (see FIG. 3B).
(3) When there are two or more extraction candidates for the target block having the same color as the adjacent block extraction color, the extraction color of the target block is determined based on the continuity of the color between the target block and the adjacent block.

  Since (1) and (2) above need not be described, (3) will be described in more detail here.

  FIG. 8 shows an example in which at least two colors {A, B} among the extracted color candidates in the target block are the same as the extracted colors of the adjacent blocks. In the figure, the size of one block is 4 × 4, but in actuality, the size is 8 × 8. Note that this is for ease of explanation.

  As shown in FIG. 8, the frequency of the extracted color A in the upper left adjacent block with respect to the target block is 3, and the frequency of the extracted color B in the upper adjacent block is 4 (these frequencies are stored in the buffer 120). Derived from extracted color and location information). Therefore, in this case, the color B having a high frequency is determined as the extracted color of the target block.

  Here, a case where the frequency of the color A of the adjacent block at the upper left of the target block in FIG. 8 is 4 and is the same as the frequency of the color B of the adjacent block will be considered. In this case, the extracted color determination unit 110 calculates the shortest distance between the pixel position having the extracted color candidate A in the target block and the pixel position of an adjacent block having the same color A extracted color. Similarly, the extraction color determination unit 110 calculates the shortest distance between the pixel position having the extraction color candidate B in the block of interest and the pixel position of an adjacent block having the same color B extraction color. Then, the shortest distance among the calculated distances is obtained, and the color that is the shortest distance is determined as the extracted color of the block of interest. In the case of FIG. 8, the distance obtained from the color B that is the extraction color candidate of the target block is the shortest, so the extraction color determination unit 110 determines the color B as the final extraction color.

  A specific example of the processing content of the extraction color determination unit 110 in the second embodiment will be described with reference to the flowchart of FIG. This figure shows a processing procedure for one block data in the extracted color determination unit 110. The processing steps substantially the same as those in FIG.

  First, in step S <b> 21, the extracted color determination unit 110 obtains a histogram of pixel colors in the target block data. When creating this histogram, R, G, and B data are counted as one set.

  Next, in step S22, the extracted color determining unit 110 determines whether there is a color having a frequency equal to or higher than a preset threshold Th from the histogram. When it is determined that there is no color having a frequency exceeding the threshold, that is, when it is determined that there is no character line drawing pixel in the target block, the process proceeds to step S23, and a control signal indicating no extracted color is multiplexed. The data is output to 180, and this processing is finished.

  If it is determined in step S22 that there is a color having a frequency equal to or higher than the threshold, that is, if it is determined that a character / line drawing pixel exists in the block of interest, the process proceeds to step S24. In step S <b> 24, the extracted color determination unit 110 outputs a signal indicating the presence of the extracted color to the multiplexing unit 180. Hereinafter, a color having a frequency equal to or higher than the threshold (a plurality of colors) may be referred to as an extraction color candidate.

  In the next step S25, the extracted color determination unit 110 determines whether or not the extracted color candidate has the same color as the extracted color of the adjacent block immediately before, directly above, upper left, and upper right of the block of interest. This is determined by referring to the data stored in 120.

  If it is determined that there is no extracted color candidate of the target block that is the same as the extracted color of the adjacent block, the most frequent color among the extracted color candidates of the target block is determined in step S26. The final extracted color is determined (see FIG. 3A).

  If it is determined that the extracted color candidates of the target block are the same as the extracted colors of the adjacent blocks, the process proceeds to step S71. In this step S71, it is determined whether two or more colors in the extracted color candidates of the block of interest match the extracted colors of the adjacent blocks. If it is determined that only one extracted color candidate matches the extracted color of the adjacent block, the process proceeds to step S72, and the extracted color candidate determined to match is determined as the extracted color of the target block (FIG. 3 ( b)).

  If it is determined in step S71 that two or more of the extracted color candidates match the extracted color of the adjacent block, the process proceeds to step S73. In this step S73, the number of pixels in the adjacent block having each color determined to match is counted. This count is determined based on the extracted color information and position information of adjacent blocks stored in the buffer 120. Next, in step S74, it is determined whether or not the number of colors that has become the most frequent is only one. If it is determined that only one color is the most frequent, in step S75, the most frequent color is determined as the extracted color of the target block (see FIG. 8).

  If the same number of colors is the most frequent, the process proceeds to step S76. In this step S76, for each color determined to be the same color as the extracted color of the adjacent block among the extracted color candidates of the target block, the minimum value of the distance between the pixel of the target block and the adjacent block is obtained. Then, the color with the shortest distance is determined as the extracted color of the block of interest (see the arrow in FIG. 8).

  When the process in any one of steps S26, S72, S75, and S76 is completed, the process proceeds to step S28. Here, the extracted color determination unit 110 outputs the RGB component data of the determined extracted color to the buffer 120, the comparison unit 140, and the encoding unit 170.

  As a result, in the input block data, the comparison unit 140 displays position information “1” at the pixel position that is the extraction color output from the extraction color determination unit 110 and “0” at the pixel position that is the non-extraction color. Can be generated. In addition, when a plurality of extracted color candidates of the target block match the extracted colors of adjacent blocks, the pixels determined as character line drawings are more likely to be the same color between blocks, and the characters of the decoded image Line art can be made more natural.

<Third Embodiment>
In the first and second embodiments, the number of pixels of the character / line drawing extracted from one block is one. In the third embodiment, an example will be described in which up to two text line drawing colors are extracted from one block.

  FIG. 9 is a block diagram of an encoding apparatus according to the third embodiment. A difference from FIG. 6 in the second embodiment is that a control signal from the extraction color determination unit 110 is also supplied to the encoding units 170 and 171. Also, the processing contents of some processing units are different. First, each processing unit that performs processing different from that of the second embodiment in the third embodiment will be described.

  The extracted color determining unit 110 determines the number of extracted colors from 0 to 2 and the extracted color, and outputs the determined color. Further, the extracted color determination unit 110 outputs the comparison unit 140, the encoding units 170 and 171 and the multiplexing unit 180 with information indicating the number of extracted colors as a control signal. Since the number of colors is three from 0 to 2, as described above, it is sufficient that the control signal has 2 bits. A control signal “00B” (B means a binary number) indicates that there is no extracted color, “01B” indicates that the number of extracted colors is 1, and “10B” indicates that the number of extracted colors is 2. However, in order to simplify the description, the number of extracted colors will be expressed in decimal numbers hereinafter.

  In the case of the control signal “0” or “1” from the extraction color determination unit 110, the processing contents of each processing unit are the same as those in the second embodiment, and thus description thereof is omitted. Therefore, the case where the control signal from the extracted color determination unit 110 is “2” will be described below. In the following, the two extracted colors determined by the extracted color determining unit 110 are expressed as E1 and E2.

  When receiving the control signal “2”, the comparison unit 140 outputs 2-bit position information for one pixel. That is, 01B is output as position information when the same pixel as the extraction color E1 exists in the input block data, and 10B is output as position information when the same pixel as the extraction color E2 exists. In other words, the comparison unit 140 sets bit 0 to “1”, bit 1 to “0” at a pixel position having the same color as the extracted color E1, and bit 0 to “0” at a pixel position having the same color as the extracted color E2. ", Position information with bit 1 set to" 1 "is generated. Further, the comparison unit 140 outputs 00B as position information when it does not match any of the colors E1 and E2. When receiving the control signal “1”, the comparison unit 140 generates position information in bit 0 and outputs the position information.

  The replacement value generation unit 150 and the selector 160 logically sum the 2-bit position information for each pixel and input and process it as 1-bit position information. Accordingly, the replacement value generation unit 150 and the selector 160 are the same as those in the first and second embodiments, except that a configuration for performing a logical sum is provided.

  When the control signal is “2”, the extracted color determining unit 110 outputs two extracted colors E1 and E2. Therefore, the encoding unit 170 performs lossless encoding of the two extracted colors E1 and E2. When the control signal is “1”, the extracted color determining unit 110 outputs only one extracted color, so the encoding unit 170 encodes one extracted color information. become.

  When the control signal is “2”, the encoding unit 171 inputs 2-bit position information for each pixel and performs lossless encoding. The type is not limited as long as it is lossless encoding. For example, two 8 × 8 1-bit planes exist, and each bit plane is losslessly encoded (run-length encoding may be used). When the control signal is “1”, the encoding unit 171 encodes the position information (1 pixel 1 bit) of the lower bit 0 of the 2 bits.

  The multiplexing unit 180 basically generates encoded data shown in FIGS. 4 (a) and 4 (b). However, the flag in the header is 2 bits (the control signal may be written as it is). When the control signal is “0”, the multiplexing unit 180 generates the encoded data string in FIG. However, when the control signal is “1”, one piece of “extracted color encoded data” is stored, and when the control signal is “2”, two pieces of extracted color information encoded data are stored.

  When the control signal is “2”, the encoded extracted color information stored in the “extracted color encoded data” is in the order of E1 and E2. Further, the extracted color E1 is assigned to bit 0 of the position information, and the extracted color E2 is assigned to bit 1 of the position information. In short, the relationship between the position information and the extracted color may be made to correspond one-to-one.

  The processing content of each processing unit in the third embodiment has been described above. Next, the process of the extraction color determination unit 110 in the third embodiment will be described in more detail.

In the third embodiment, the extraction color determination unit 110 performs processing similar to the first and second embodiments so far by creating a histogram relating to the color of the block of interest and setting a preset threshold value (first Extraction color candidates are obtained for colors having a frequency equal to or greater than (threshold). Then, the extracted color determining unit 110 determines the extracted color and the number according to the following conditions.
(1) If the number of extracted color candidates is one, that color is determined as an extracted color candidate.
(2) If there is one extracted color candidate that is the same as the extracted color of the adjacent block, that color is determined as the extracted color candidate.
(3) When there are two or more same colors as the extracted colors of adjacent blocks in the extracted color candidates, the total number of the first and second frequencies is set to a preset threshold value (second threshold value). If it exceeds, the two colors are determined as the extracted colors of the block of interest. When the total number is equal to or smaller than the second threshold, only the first extracted color candidate is determined as the extracted color of the target block.

  The above (1) and (2) will not be described. Therefore, here, (3) will be described.

  FIG. 10 shows an example in which the colors A and D are the same as the extracted colors of the adjacent blocks among the extracted color candidates having a frequency equal to or higher than the first threshold in the target block. In the figure, the size of one block is 4 × 4, but in actuality, the size is 8 × 8. Note that this is for ease of explanation.

  Now, in the case shown in FIG. 10, the extracted color determination unit 110 calculates the sum of the frequencies of the colors A and D in the target block, and whether the total value exceeds a preset threshold (second threshold Th2). Judge whether or not. When it is determined that the total value exceeds the second threshold value, two colors A and D are determined as the extracted colors of the block of interest. If the total value is less than or equal to the second threshold value, one color of the first color D is determined as the extracted color of the block of interest.

  A specific example of the processing content of the extraction color determination unit 110 in the third embodiment will be described with reference to the flowchart of FIG. This figure shows a processing procedure for one block data in the extracted color determination unit 110. The processing steps substantially the same as those in FIG.

  First, in step S <b> 21, the extracted color determination unit 110 obtains a histogram of pixel colors in the target block data. When creating this histogram, R, G, and B data are counted as one set.

  Next, in step S22, the extracted color determination unit 110 determines whether there is a color having a frequency equal to or higher than a preset first threshold value Th1 from the histogram. If it is determined that there is no color having a frequency exceeding the threshold, that is, if it is determined that there is no character line drawing pixel in the block of interest, the process proceeds to step S23, and no extracted color (the number of extracted colors is “0”). ) Is output to the multiplexing unit 180, and the present process ends.

  If there is a color having a frequency equal to or higher than the first threshold Th1, it is set as an extraction color candidate. In step S1001, it is determined whether the number of extracted color candidates is one. If there is one extracted color candidate, the process proceeds to step S1005, and a control signal indicating that the number of extracted colors is “1” is output to the multiplexing unit 180. In step S1006, the color is determined as the extracted color of the target block, and in step S1009, RGB component data of the determined extracted color is output to the buffer 120, the comparison unit 140, and the encoding unit 170.

  If it is determined that there are a plurality of extracted color candidates, the process advances to step S1002. In step S1002, it is determined whether or not one extraction candidate has the same color as the extraction color of the adjacent block. If it is determined that only one of the extraction candidates has the same color as the adjacent block, the processing of steps S1005, S1006, and S1109 is performed. Note that. The extracted color determined in step S1006 is determined to be the same color as the adjacent block in the extraction candidate.

  If there are two or more extraction candidates that have the same color as the adjacent block, the process of step S1003 is performed. In step S1003, a total value of the frequencies of the top two colors determined to be the same color as the adjacent block among the extracted color candidates is calculated. For example, it is assumed that there are three extraction color candidates of the target block {C1, C2, C3}, of which the colors C1 and C3 are extraction colors of adjacent blocks. In this case, the top two frequencies of matching colors are the colors C1 and C3. Therefore, the total value of the frequencies of the colors C1 and C3 is calculated.

  In step S1004, it is determined whether the calculated total value exceeds a preset second threshold Th2. If “total value ≦ second threshold Th2”, the processes of steps S1005, S1006, and S1009 are performed. Note that. The extracted color determined in step S1006 is the color of the first extracted color candidate (color C1 in the above example) used when calculating the total value.

  If it is determined in step S1004 that “total value> second threshold Th2”, the process proceeds to step S1007 to output a control signal indicating that the number of extracted colors is “2”. In step S1008, two colors (in the above example, colors C1 and C3) are determined as the extracted colors of the block of interest, and in step S1009, RGB component data of the determined two extracted colors are obtained. , Output to the buffer 120, the comparison unit 140, and the encoding unit 170.

  As described above, in addition to the functions and effects of the first and second embodiments, it is possible to deal with the case where a pixel constituting a character line image exists in the block of interest and there are two colors. Become.

<Modification of Third Embodiment>
In the third embodiment, when a pixel of a two-color character / line image is found in the block of interest, the encoding unit 171 encodes 2-bit position information per pixel. The encoding unit 171 in the embodiment uses run-length encoding. Therefore, when encoding 2-bit position information per pixel, the binary plane of bit 0 and the binary plane of bit 1 are encoded separately. Here, when the run length encoding is performed by raster scanning the bit plane, and the bit value (0 or 1) changes frequently, the compression rate is expected to decrease. In other words, it can be said that the compression rate of the position information can be increased by reducing the number of times the bit value is switched when scanning the bit plane.

  Therefore, in the modification of the third embodiment, a modification of color selection in the extraction color determination unit 110 for increasing the compression rate of position information will be described.

In the modified example of the third embodiment, the extraction color determination unit 110 performs processing similar to the first and second embodiments so far by creating a histogram relating to the color of the block of interest and setting a preset threshold value. An extraction color candidate is obtained for a color having a frequency equal to or greater than (first threshold). Then, the extracted color determining unit 110 determines the extracted color and the number according to the following conditions.
(1) If there is no extracted color candidate, the number of extracted colors is set to “0”.
(2) If there is only one extracted color candidate, that color is determined as the extracted color, and the number of extracted colors is also set to “1”.
(3) When there are a plurality of extraction color candidates, the color with the first frequency is set as the extraction color (first extraction color). The following are sub-conditions in the condition (3).
(3a) Among the extracted color candidates whose frequency is second or higher, a search is made for a match with the extracted color of the adjacent block. If there is no matching color, it is determined that the number of extracted colors of the target block is one.
(3b) If there is only one extracted color candidate whose frequency is second or higher that matches the extracted color of the adjacent block, that color is determined as the extracted color (second extracted color); It is determined that the number of extracted colors is two.
(3c) In the case where there are a plurality of extracted color candidates whose frequency is the second or higher that matches the extracted color of the adjacent block, the color with the highest continuity of pixel positions when raster scanning of each color is performed Is determined as the extracted color (second extracted color), and the number of extracted colors is determined to be two.

  Here, the sub condition (3c) will be described with reference to FIG. FIG. 12 shows an example in which the extraction color candidate {E, D, A} (in order of frequency) of the block of interest is found. Since the frequency of the color E is the highest, the first extracted color is the color E. Both the colors D and A with the second highest frequency match the color of the adjacent block. Accordingly, one of the colors D and A is determined as the second extracted color.

  Focusing on color A, if the target block is raster scanned, it becomes 1, 0, 0, ..., 0, 1, 0, 0, ... 0, 1 and the number of changes from 1 to 0 and from 0 to 1 is 4 times. On the other hand, in the case of the color D, when raster scanning is performed, the number of changes is 2, such as 0, 0, 0, 0, 1, 1, 1, 1, 0, 0,. Therefore, it can be said that the color D has a smaller number of changes than the color A, and the encoding efficiency by the run-length encoding is high. Therefore, in this case, the colors E and D are determined as the extracted colors of the target block, and the number of extracted colors is determined as “2”.

<Other embodiments>
You may implement | achieve the function of the apparatus of each embodiment described above and the modified example with the computer program which makes a computer read and execute. In this case, the part corresponding to each processing unit can be realized as a part program (subroutine or function).

  Also, the computer program is usually stored in a computer-readable storage medium such as a CD-ROM, and is set in a reading unit (CD-ROM drive or the like) provided in the computer and copied or installed in the system. It becomes possible. Therefore, it is obvious that such a computer-readable storage medium falls within the scope of the present invention.

It is a block block diagram of the image coding apparatus in 1st Embodiment. It is a flowchart for showing the processing content of the extraction color determination part in 1st Embodiment. It is a figure which shows the specific example of a process of the extraction color determination part in 1st Embodiment. It is a figure which shows the data structure of the coding data for 1 block which the image coding apparatus in 1st Embodiment produces | generates. It is a block block diagram of the decoding apparatus in 1st Embodiment. It is a block block diagram of the image coding apparatus in 2nd Embodiment. It is a flowchart for showing the processing content of the extraction color determination part in 2nd Embodiment. It is a figure which shows the specific example of the process of the extraction color determination part in 2nd Embodiment. It is a block block diagram of the image coding apparatus in 3rd Embodiment. It is a figure which shows the specific example of the process of the extraction color determination part in 3rd Embodiment. It is a flowchart for showing the processing content of the extraction color determination part in 3rd Embodiment. It is a figure which shows the specific example of a process of the extraction color determination part of the modification of 3rd Embodiment.

Claims (11)

An image encoding device for encoding multi-value image data,
Input means for inputting image data in block units to be subjected to lossy encoding from multi-valued image data to be encoded;
Extraction means for extracting color information indicating the color of the pixel of the line drawing in the image data of the input block;
Position information generating means for generating position information for identifying the position of the pixel of the line drawing in the target block and the position of the pixel of the non-line drawing according to the color information extracted by the extracting means;
Lossless encoding means for losslessly encoding the position information generated by the position information generation means and the color information extracted by the extraction means;
In order to replace the value of a line drawing pixel according to the position information generated by the position information generation means, a replacement value calculation means for calculating an average value of the non-line drawing pixel value as a replacement value;
Replacement means for replacing the value of the pixel of the line drawing in the block of interest with the replacement value;
Irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplex means for multiplexing and outputting encoded data of position information and color information generated by the lossless encoding means, and encoded data of image data generated by the lossy encoding means,
The extraction means includes
Storage means for storing color information extracted from already encoded blocks;
Candidate determination means for determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination unit that refers to the storage unit and determines whether there is a color that matches color information of an adjacent block among the extracted color candidates determined by the candidate determination unit;
Selection means for preferentially selecting color information determined by the determination means in the extracted color candidates as color information of a line drawing in the block of interest;
When the determination unit determines that there is no color that matches the color information of an adjacent block among the extraction color candidates determined by the candidate determination unit, the selection unit selects the extraction color candidate having the maximum frequency as the target block. Select as the color information of the line drawing in
When the determining means determines that there is a color that matches the color information of the adjacent block among the extracted color candidates determined by the candidate determining means, the selecting means , color information number of occurrences due to the combination of the extracted color candidates in the target block and the color information extracted by the adjacent blocks is maximized, picture image encoding apparatus and selecting as the color information of the line drawing in the target block. An image encoding device for encoding multi-value image data,
Input means for inputting image data in block units to be subjected to lossy encoding from multi-valued image data to be encoded;
Extraction means for extracting color information indicating the color of the pixel of the line drawing in the image data of the input block;
Position information generating means for generating position information for identifying the position of the pixel of the line drawing in the target block and the position of the pixel of the non-line drawing according to the color information extracted by the extracting means;
Lossless encoding means for losslessly encoding the position information generated by the position information generation means and the color information extracted by the extraction means;
In order to replace the value of a line drawing pixel according to the position information generated by the position information generation means, a replacement value calculation means for calculating an average value of the non-line drawing pixel value as a replacement value;
Replacement means for replacing the value of the pixel of the line drawing in the block of interest with the replacement value;
Irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplex means for multiplexing and outputting encoded data of position information and color information generated by the lossless encoding means, and encoded data of image data generated by the lossy encoding means,
The extraction means includes
Storage means for storing color information extracted from already encoded blocks;
Candidate determination means for determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination unit that refers to the storage unit and determines whether there is a color that matches color information of an adjacent block among the extracted color candidates determined by the candidate determination unit;
Selection means for preferentially selecting color information determined by the determination means in the extracted color candidates as color information of a line drawing in the block of interest;
When the determination unit determines that there is no color that matches the color information of an adjacent block among the extraction color candidates determined by the candidate determination unit, the selection unit selects the extraction color candidate having the maximum frequency as the target block. Select as the color information of the line drawing in
The storage means further stores the position information generated based on an already encoded block,
When the determination unit determines that there are two or more colors that match the color information of adjacent blocks among the extracted color candidates determined by the candidate determination unit, the selection unit selects the extracted color of the adjacent block and When there is one color that matches the color of the maximum frequency in the adjacent block among the extracted color candidates of the target block determined to match, the color is selected as the color information of the line drawing in the target block. picture image encoding apparatus for. The determination unit determines that there are two or more colors that match the color information of the adjacent block among the extracted color candidates determined by the candidate determination unit, and the attention is determined to match the extracted color of the adjacent block If it is determined that there are multiple colors that match the maximum frequency color in the adjacent block among the color candidates for the block,
The selection unit selects the color information having the shortest distance as the color information of the line drawing in the target block based on the color information and position information extracted in the adjacent block and the color and position of the extracted color candidate in the target block. The image coding apparatus according to claim 2 , wherein: An image encoding device for encoding multi-value image data,
Input means for inputting image data in block units to be subjected to lossy encoding from multi-valued image data to be encoded;
Extraction means for extracting color information indicating the color of the pixel of the line drawing in the image data of the input block;
Position information generating means for generating position information for identifying the position of the pixel of the line drawing in the target block and the position of the pixel of the non-line drawing according to the color information extracted by the extracting means;
Lossless encoding means for losslessly encoding the position information generated by the position information generation means and the color information extracted by the extraction means;
In order to replace the value of a line drawing pixel according to the position information generated by the position information generation means, a replacement value calculation means for calculating an average value of the non-line drawing pixel value as a replacement value;
Replacement means for replacing the value of the pixel of the line drawing in the block of interest with the replacement value;
Irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplex means for multiplexing and outputting encoded data of position information and color information generated by the lossless encoding means, and encoded data of image data generated by the lossy encoding means,
The extraction means includes
Storage means for storing color information extracted from already encoded blocks;
Candidate determination means for determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination unit that refers to the storage unit and determines whether there is a color that matches color information of an adjacent block among the extracted color candidates determined by the candidate determination unit;
Selection means for preferentially selecting color information determined by the determination means in the extracted color candidates as color information of a line drawing in the block of interest;
When the determination unit determines that the extracted color candidate has one color that matches the color information of the adjacent block, the selection unit selects the color determined to match as the color information of the target block And
The extracting means further determines the frequency among the extracted color candidates determined to match when the determining means determines that there are a plurality of colors that match the color information of adjacent blocks in the extracted color candidates. A second determining means for determining whether or not the total number of the first and second colors exceeds a preset threshold;
When the second determining means determines that the total number exceeds the preset threshold, the selecting means displays the two-color information with the first and second frequencies as the line drawing of the target block. picture image encoding apparatus and selecting as the color information. An image encoding device for encoding multi-value image data,
Input means for inputting image data in block units to be subjected to lossy encoding from multi-valued image data to be encoded;
Extraction means for extracting color information indicating the color of the pixel of the line drawing in the image data of the input block;
Position information generating means for generating position information for identifying the position of the pixel of the line drawing in the target block and the position of the pixel of the non-line drawing according to the color information extracted by the extracting means;
Lossless encoding means for losslessly encoding the position information generated by the position information generation means and the color information extracted by the extraction means;
In order to replace the value of a line drawing pixel according to the position information generated by the position information generation means, a replacement value calculation means for calculating an average value of the non-line drawing pixel value as a replacement value;
Replacement means for replacing the value of the pixel of the line drawing in the block of interest with the replacement value;
Irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplex means for multiplexing and outputting encoded data of position information and color information generated by the lossless encoding means, and encoded data of image data generated by the lossy encoding means,
The extraction means includes
Storage means for storing color information extracted from already encoded blocks;
Candidate determination means for determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination unit that refers to the storage unit and determines whether there is a color that matches color information of an adjacent block among the extracted color candidates determined by the candidate determination unit;
Selection means for preferentially selecting color information determined by the determination means in the extracted color candidates as color information of a line drawing in the block of interest;
When the determination unit determines that the extracted color candidate has one color that matches the color information of the adjacent block, the selection unit selects the color determined to match as the color information of the target block And
When the determining unit determines that the extracted color candidates include a plurality of colors that match the color information of adjacent blocks, the frequency of the extracted color candidates determined to match is the first. Is selected as the first color information of the line drawing of the target block, and the color information that maximizes the encoding efficiency by the lossless encoding means among the extracted color candidates in the second and subsequent positions is selected. picture image encoding apparatus and selecting as the second color information of the line drawing. A control method for an image encoding device for encoding multi-value image data,
An input step in which input means inputs image data in units of blocks to be subjected to lossy encoding from multi-valued image data to be encoded;
An extraction step in which the extraction means extracts color information indicating the color of the pixel of the line drawing in the image data of the input block;
A position information generating step for generating position information for identifying a position of a pixel of a line drawing and a position of a pixel of a non-line drawing in the target block according to the color information extracted in the extraction step;
A lossless encoding means for losslessly encoding the position information generated in the position information generation step and the color information extracted in the extraction step;
A replacement value calculating unit that calculates an average value of non-line drawing pixel values as a replacement value in order to replace the pixel value of the line drawing according to the position information generated in the position information generation step;
A replacement step in which the replacement means replaces the value of the pixel of the line drawing in the target block with the replacement value;
An irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplexing means for multiplexing and outputting the encoded data of the position information and color information generated by the lossless encoding process and the encoded data of the image data generated by the lossy encoding process A process,
The extraction step includes
A storage step of storing in the memory the color information extracted from the already encoded blocks;
A candidate determination step of determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination step of referring to the memory and determining whether there is a color that matches the color information of an adjacent block among the extracted color candidates determined in the candidate determination step;
A selection step of preferentially selecting the color information determined to match in the determination step in the extracted color candidates as the color information of the line drawing in the block of interest,
When the determination step determines that there is no color that matches the color information of an adjacent block among the extraction color candidates determined in the candidate determination step, the selection step selects an extraction color candidate having the maximum frequency as the target block Select as the color information of the line drawing in
If the determination step determines that there is a color that matches the color information of an adjacent block among the extraction color candidates determined in the candidate determination step, the selection step includes: Control of an image coding apparatus, wherein color information that maximizes the number of appearances by a combination of color information extracted in adjacent blocks and extracted color candidates in the target block is selected as color information of a line drawing in the target block Method. A control method for an image encoding device for encoding multi-value image data,
An input step in which input means inputs image data in units of blocks to be subjected to lossy encoding from multi-valued image data to be encoded;
An extraction step in which the extraction means extracts color information indicating the color of the pixel of the line drawing in the image data of the input block;
A position information generating step for generating position information for identifying a position of a pixel of a line drawing and a position of a pixel of a non-line drawing in the target block according to the color information extracted in the extraction step;
A lossless encoding means for losslessly encoding the position information generated in the position information generation step and the color information extracted in the extraction step;
A replacement value calculating unit that calculates an average value of non-line drawing pixel values as a replacement value in order to replace the pixel value of the line drawing according to the position information generated in the position information generation step;
A replacement step in which the replacement means replaces the value of the pixel of the line drawing in the target block with the replacement value;
An irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplexing means for multiplexing and outputting the encoded data of the position information and color information generated by the lossless encoding process and the encoded data of the image data generated by the lossy encoding process A process,
The extraction step includes
A storage step of storing in the memory the color information extracted from the already encoded blocks;
A candidate determination step of determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination step of referring to the memory and determining whether or not there is a color that matches color information of an adjacent block among the extracted color candidates determined in the candidate determination step;
A selection step of preferentially selecting the color information determined to match in the determination step in the extracted color candidates as the color information of the line drawing in the block of interest,
When the determination step determines that there is no color that matches the color information of an adjacent block among the extraction color candidates determined in the candidate determination step, the selection step selects an extraction color candidate having the maximum frequency as the target block Select as the color information of the line drawing in
The storing step further stores the position information generated based on the already encoded block in the memory,
When the determination step determines that there are two or more colors that match the color information of adjacent blocks among the extraction color candidates determined in the candidate determination step, the selection step includes the extraction color of the adjacent block and When there is one color that matches the color of the maximum frequency in the adjacent block among the extracted color candidates of the target block determined to match, that color is selected as the color information of the line drawing in the target block.
A control method for an image encoding device. The determination that the determination step determines that there are two or more colors that match the color information of adjacent blocks in the extracted color candidates determined in the candidate determination step, and the attention is determined to match the extraction color of the adjacent blocks If it is determined that there are multiple colors that match the maximum frequency color in the adjacent block among the color candidates for the block,
The selection step selects the color information having the shortest distance as the color information of the line drawing in the target block based on the color information and position information extracted in the adjacent block and the color and position of the extracted color candidate in the target block. The method of controlling an image encoding device according to claim 7. A control method for an image encoding device for encoding multi-value image data,
An input step in which input means inputs image data in units of blocks to be subjected to lossy encoding from multi-valued image data to be encoded;
An extraction step in which the extraction means extracts color information indicating the color of the pixel of the line drawing in the image data of the input block;
A position information generating step for generating position information for identifying a position of a pixel of a line drawing and a position of a pixel of a non-line drawing in the target block according to the color information extracted in the extraction step;
A lossless encoding means for losslessly encoding the position information generated in the position information generation step and the color information extracted in the extraction step;
A replacement value calculating unit that calculates an average value of non-line drawing pixel values as a replacement value in order to replace the pixel value of the line drawing according to the position information generated in the position information generation step;
A replacement step in which the replacement means replaces the value of the pixel of the line drawing in the target block with the replacement value;
An irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplexing means for multiplexing and outputting the encoded data of the position information and color information generated by the lossless encoding process and the encoded data of the image data generated by the lossy encoding process A process,
The extraction step includes
A storage step of storing in the memory the color information extracted from the already encoded blocks;
A candidate determination step of determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination step of referring to the memory and determining whether or not there is a color that matches color information of an adjacent block among the extracted color candidates determined in the candidate determination step;
A selection step of preferentially selecting the color information determined to match in the determination step in the extracted color candidates as the color information of the line drawing in the block of interest,
If it is determined in the determination step that there is one color that matches the color information of the adjacent block among the extracted color candidates, the selection step selects the color determined to match as the color information of the target block And
In the extraction step, when it is determined in the determination step that there are a plurality of colors that match the color information of adjacent blocks in the extraction color candidates, the frequency among the extraction color candidates determined to match is determined. A second determination step of determining whether or not the total number of the first and second colors exceeds a preset threshold;
When it is determined in the second determination step that the total number exceeds the preset threshold value, the selection step displays the two-color information with the first frequency and the second frequency as the line drawing of the target block. Select as color information
A control method for an image encoding device. A control method for an image encoding device for encoding multi-value image data,
An input step in which input means inputs image data in units of blocks to be subjected to lossy encoding from multi-valued image data to be encoded;
An extraction step in which the extraction means extracts color information indicating the color of the pixel of the line drawing in the image data of the input block;
A position information generating step for generating position information for identifying a position of a pixel of a line drawing and a position of a pixel of a non-line drawing in the target block according to the color information extracted in the extraction step;
A lossless encoding means for losslessly encoding the position information generated in the position information generation step and the color information extracted in the extraction step;
A replacement value calculating unit that calculates an average value of non-line drawing pixel values as a replacement value in order to replace the pixel value of the line drawing according to the position information generated in the position information generation step;
A replacement step in which the replacement means replaces the value of the pixel of the line drawing in the target block with the replacement value;
An irreversible encoding means for irreversibly encoding the image data of the replaced block;
Multiplexing means for multiplexing and outputting the encoded data of the position information and color information generated by the lossless encoding process and the encoded data of the image data generated by the lossy encoding process A process,
The extraction step includes
A storage step of storing in the memory the color information extracted from the already encoded blocks;
A candidate determination step of determining the frequency of each color in the block of interest and determining a color having a frequency equal to or higher than a preset threshold as an extraction color candidate;
A determination step of referring to the memory and determining whether or not there is a color that matches color information of an adjacent block among the extracted color candidates determined in the candidate determination step;
A selection step of preferentially selecting the color information determined to match in the determination step in the extracted color candidates as the color information of the line drawing in the block of interest,
If it is determined in the determination step that there is one color that matches the color information of the adjacent block among the extracted color candidates, the selection step selects the color determined to match as the color information of the target block And
In the selection step, when it is determined in the determination step that there are a plurality of colors that match the color information of adjacent blocks in the extracted color candidates, the frequency of the extracted color candidates determined to match is first. Is selected as the first color information of the line drawing of the target block, and the color information that maximizes the encoding efficiency by the lossless encoding process among the extracted color candidates after the second position is selected. Select as second color information of line drawing
A control method for an image encoding device. By executing Mase read into the computer, the computer, causes the computer to function as the image coding apparatus according to any one of claims 1 to 5.

Images