JP2008311792A - Device for encoding image and control method therefor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To further improve encoding efficiency, by determining the threshold as to whether a palletization is conducted, according to block sizes of encoding units. <P>SOLUTION: A blocking section 200 inputs an image data while using the set block size as a unit and stores the image data in a buffer memory 201. A histogram preparation section 202 prepares a histogram for the colors of the image data in one block section. A decision section 203 determines the threshold from the set block size, compares a color number appearing in the prepared histogram and the threshold and outputs the result of the comparison to a displacement section 204. The decision section 203 prepares a conversion table that allocates pallet numbers, regarding each color appearing in the block, when the a color number is lower than the threshold, and sets it to the displacement section 204. The displacement section 204 reads the image data corresponding to a single block section from the buffer memory and outputs the image data to an encoding section 205. However, when the conversion table is set from the decision section 203, a pallet conversion is conducted, and the result is output to an encoding section 205 as a pixel data. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

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

  Conventionally, as an irreversible compression method for still images, a JPEG method using discrete cosine transform and a method using Wavelet transform are often used. Also, as a lossless compression method, a Runlength method, a method using predictive coding, and the like are widely known.

  As reversible compression methods, Pakcbits, international standardized methods such as JPEG-LS and JPEG2000 encoding methods have been used. In a lossless encoding method using predictive encoding, in order to increase encoding efficiency, it is necessary to increase prediction accuracy. For example, when the range to be referred to for prediction is widened or the frequency of appearance is recorded, the prediction accuracy can be improved and the coding efficiency can be increased.

Furthermore, as a method for reducing the prediction error, a method has been proposed in which appearance pixel values are collected for an image with sparse appearance pixel values to suppress the prediction error (for example, Patent Document 1).
JP 2005-348390 A

  In Patent Document 1, it is possible to vary the number of application levels for grouping appearance pixels. However, depending on the number of application levels, the encoding efficiency may deteriorate, and there is still room for improvement. is there.

  The present invention has been made in view of such a problem, and intends to provide a technique for further improving the coding efficiency by determining a threshold value for determining whether or not to palletize according to the block size. is there.

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 that encodes image data and generates encoded data,
Setting means for setting the block size of the encoding unit;
Determining means for determining a threshold according to the set block size;
Input means for inputting image data in units of blocks of a set size;
A histogram creation means for creating a histogram of the color of the input block of interest;
Comparing means for calculating the number of colors appearing in the block of interest from the created histogram, comparing the threshold value determined by the determining means with the number of colors, and determining whether the number of colors is equal to or less than the threshold value; ,
A conversion means for assigning a palette number to each color appearing in the block of interest and converting each pixel data in the block of interest into a palette number when the comparison means determines that the number of colors is equal to or less than the threshold;
When the comparison unit determines that the number of colors is equal to or less than the threshold value, the palette number converted by the conversion unit is losslessly encoded as pixel data, and the code obtained by encoding together with information indicating palette conversion. Output data,
If the comparison means determines that the number of colors is greater than the threshold value, each pixel data constituting the block of interest is losslessly encoded and the encoded data obtained by encoding is output together with information indicating non-palette conversion. Encoding means.

  According to the present invention, it is possible to further improve the encoding efficiency by determining a threshold value for determining whether or not to perform palletization according to the block size of the encoding unit.

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

  In the following, an example in which the image encoding technique is applied to a printing apparatus will be described. However, the image encoding apparatus alone or an apparatus installed in another information processing apparatus may be used.

<First Embodiment>
FIG. 2 is a system schematic diagram according to the present embodiment. In the figure, 10 is a printing apparatus, and 50 is a host computer that outputs print data. The printing apparatus 10 includes a printer control unit 11 that receives and renders print data, a large-capacity storage device 105 such as a hard disk, and a printer engine unit 110 that prints a visible image on recording paper. The printer control unit 11 includes various processing circuits, including a CPU, a program executed by the CPU, and a memory used as a work area. Further, the host computer 50 and the printing apparatus 10 are connected via a network, but other interfaces (for example, USB) may be used and the type thereof is not limited. In this embodiment, the printer engine unit 110 is described as adopting an electrophotographic system, but may be ink jet, thermal recording, or the like, and the type thereof is not limited. The host computer 50 is a personal computer, for example. For example, a document editing application program is executed on the host computer 50. Assume that the user gives a print instruction using the keyboard or mouse of the host computer 50. At this time, a printer driver installed in the OS (operating system) of the host computer 50 inputs data to be printed from the application, and generates print data described in a page description language (PDL) that can be interpreted by the printing apparatus 10. . Then, the print data is transmitted to the printing apparatus 10 via the OS. The printer control unit 11 of the printing apparatus 10 interprets the print data received from the host computer 50, renders the data in units of pages, stores it in the storage device 105, and sends the rendering result stored in the storage device 105 to the printer engine unit 110. Output. The printing apparatus 10 can also receive print data from other host computers.

  FIG. 1 is a block diagram of the printer control unit 11 of the printing apparatus 10 according to the embodiment.

  In the figure, reference numeral 100 denotes a rendering unit that interprets print data (PDL data) received from the host computer 50 and generates a bitmap image. The rendering unit 100 includes a drawing memory for one page and an attribute flag memory for storing attributes for each pixel. For the character code specified by the PDL data, data is read from a font memory (not shown) according to the specified font type, a character pattern is generated, and developed into a drawing memory. Also, drawing processing of line drawings and various figures designated by PDL data is performed. Further, when image data such as a natural image is included in the PDL data, the image data is expanded. In addition, when the rendering unit 100 performs rendering processing on the drawing memory, the attribute flag (character / line image, natural image, information identifying CG, information indicating chromatic / achromatic color, halftone dot) for each corresponding pixel position. (Attribute of information indicating whether or not) is stored in the attribute flag memory. This attribute flag memory stores and manages an attribute flag in units of 1 byte (8 bits) for one pixel in the drawing memory. Each of the above-described attributes is represented by using some of the 8 bits.

  Reference numeral 101 denotes a buffer memory for compression processing, 102 denotes an attribute flag encoding unit, 103 denotes a color image encoding unit, 104 denotes a buffer memory for writing, and 105 denotes a large-capacity storage device such as a hard disk device. 106 is a buffer for decoding processing, 107 is an attribute flag decoding unit, 108 is a color image decoding unit, 109 is a decoded attribute flag, and a buffer memory for developing image data. An image processing unit 110 performs color conversion (conversion to the recording color component CMYK) and gradation correction for the decoded image data with reference to attribute data, and outputs the obtained data to the printer engine 110.

  In the above configuration, it is assumed that print data (PDL data) is received via an interface (not shown). In this case, the rendering unit 100 interprets the print data, draws (renders) image data in units of pages, and stores the attribute flag at the pixel position corresponding to the attribute flag memory at that time. In rendering, each pixel is an RGB color space, and one component is 8 bits. This color space and the number of bits are merely examples, and other color spaces may be used.

  When the storage of the image data for one page and the attribute flag is completed, the rendering unit 100 outputs the data to the buffer memory 101 and starts the rendering process for the next page.

  The image encoding processing unit 103 encodes the image data stored in the buffer memory, and the attribute flag encoding unit 102 encodes the attribute flag stored in the buffer memory. The generated encoded image data and encoding attribute flag are stored as files in the storage device 105 via the buffer memory 104, respectively.

  On the other hand, the attribute flag decoding unit 107 and the image decoding processing unit 108 buffer the encoded image data file and the encoded attribute flag file stored in the storage device 105 in synchronization with the print output process of the printer engine unit 111. Read into the memory 106. Then, each performs decoding processing and stores the decoding result in the buffer memory 109. The image processing unit 110 refers to the decoded attribute flag, converts the decoded image data (RGB data) into a YMCK color space according to the attribute, performs correction processing, and performs printing processing of the printer engine unit 111. A process of outputting in synchronization is performed.

  As a result, the image and attribute flag encoding process and the decoding process can be made asynchronous with each other through the storage device 105. Therefore, the rendering unit 100, the image encoding unit 103, and the attribute flag encoding unit 102 can encode the image and attribute flag of each page regardless of the printing speed of the printer engine unit 111. As a result, the host computer 50 is quickly released from the print data output process to the printing apparatus 10 of the present embodiment. In addition, the printing apparatus can cope with reception of print data from another computer (not shown).

  Next, the encoding process of the control unit 150 in the printing apparatus 10 of the present embodiment having the above-described configuration will be described below with reference to the flowchart of FIG.

  When a print job (PDL data) is received from the host computer 50, the rendering unit 100 first performs PDL analysis in step S11. In step S12, the rendering unit 100 generates a bitmap image by rendering based on the PDL. In step S13, the rendering unit 100 obtains attribute information for each pixel during PDL analysis, generates an attribute flag, and stores the attribute flag in the attribute memory. Note that rendering processing and generation and storage of attribute flags are performed in parallel.

  As described above, the rendering unit 100 analyzes the print data described in PDL received from the host computer 50, interprets the PDL command included therein, and generates an image (bitmap image) for printing. . At this time, referring to information indicating the type of command in the PDL command, the attribute of each pixel is determined and stored in the attribute flag memory as an attribute flag.

  When the rendering for one page is completed, the rendered bitmap image is encoded by the image encoding processing unit 103 according to lossless compression encoding (in this embodiment, JPEG-LS is adopted) (step S14). . On the other hand, the attribute flag data is reversibly compressed by the encoder 102 (step S15). As the encoding of the attribute flag data, run-length encoding is used, but other encoding techniques (for example, predictive encoding, packbits encoding, etc.) may be used. The encoded data generated by each is stored in the storage device 105 via the buffer 104 (step S16).

  In the above process, the attribute flag encoding unit 102 only performs raster scanning and run-length encoding of the attribute flag for one page, and a description thereof will be omitted. Hereinafter, the image encoding processing unit 103 in the embodiment will be described in more detail.

  FIG. 4 is a detailed block diagram of the image encoding processing unit 103 in the embodiment.

  The image encoding device unit 103 includes a blocking unit 200, a buffer memory 201, a histogram creation unit 202, a determination unit 203, a replacement unit 204, and an encoding unit 205, as illustrated. The processing contents of the image encoding processing unit 103 according to the embodiment will be described below with reference to FIG.

  In accordance with an instruction from the operation unit 151 or control information stored in the print data, the control unit 150 generates a control signal for specifying a block size (m × n pixels m and n) that is an encoding unit of image data. Generate. Then, the control unit 150 outputs the generated control signal to the image encoding processing unit 103 prior to the encoding process for one page. Note that the block size specified by the operation unit 151 may differ from the block size specified by the control information included in the print data. If the print data includes a command for specifying the block size, the control unit 150 determines the block size of the encoding unit according to the command. If there is no command for specifying the block size in the print data, the size instructed from the operation unit 151 is determined as the block size of the encoding unit.

  The blocking unit 200 reads image data (hereinafter referred to as block data) of a block (m × n pixels) having a size according to the control signal from the control unit 150 from the buffer memory 101 (see FIG. 1). Store in the buffer 201. The order of the blocks read by the blocking unit 200 is the raster scan order in units of blocks.

  The histogram creation unit 202 creates histogram information (appearance frequency for each color) of block data stored in the buffer memory 201, and outputs the created histogram information to the determination unit 203.

The determination unit 203 has a built-in lookup table as a threshold storage unit, and reads the threshold Th from the lookup table based on the block size indicated by the control information from the control unit 150. Further, the determination unit 203 calculates the number Nc of colors that appear in the block of interest based on the histogram information from the histogram creation unit 202. Then, the determination unit 203 compares the calculated color number Nc with the threshold Th, determines whether the color number Nc is equal to or less than the threshold Th, and performs the following processing according to the determination result.
-When the number of colors Nc is less than or equal to the threshold value;
The determination unit 203 outputs, to the replacement unit 204 and the encoding unit 205, a signal for validating the replacement process (also a signal indicating that the number of colors ≦ threshold Th). Then, the determination unit 203 rearranges the colors in descending order of frequency of appearance in the block of interest. Then, the determination unit 203 makes a pair of the rearranged color data (RGB values) and the palette numbers 0, 1, 2,..., Nc−1 as shown in FIG. A palette conversion table) is created and output to the replacement unit 204 and the encoding unit 205.
When the number of colors Nc is greater than the threshold;
The determination unit 203 outputs to the replacement unit 204 a signal that invalidates the replacement process (also a signal indicating the number of colors> threshold Th).

  The replacement unit 204 reads block data from the buffer memory 201 for each block size indicated by the control signal from the control unit 150, performs replacement processing, and outputs the block data to the encoding unit 205. More specifically, it is as follows.

  When the signal from the determination unit 203 indicates that the replacement process is invalid, the replacement unit 203 performs raster scan within the block of interest stored in the buffer memory 201, reads each pixel data, and codes the read pixel data as it is. To the conversion unit 205.

  On the other hand, when the signal from the determination unit 203 indicates that the replacement process is valid, the replacement unit 204 stores the palette conversion table from the determination unit 203 as a lookup table in an internal writable memory. Then, the replacement unit 204 reads each pixel data by performing a raster scan of the block of interest stored in the buffer 201, and converts the read pixel data into a palette number using a lookup table. Also, the replacement unit 204 outputs the converted data (pallet number) to the encoding unit 205 as data of one component of RGB (in the embodiment, R component). Further, the replacement unit 204 outputs “0” as a dummy to the bus for other components (G, B components).

  Next, the encoding unit 205 will be described.

  The encoding unit 205 generates and outputs a file header prior to encoding one image. In this file header, the size of the image data to be encoded (number of pixels in the horizontal and vertical directions), the type of color space, the number of bits of each component, information specifying the block size, etc. are necessary for the decoding process. Information is stored. The encoding unit 205 performs the following processing when encoding in block units.

  When receiving the control signal indicating that the replacement process is invalid from the determination unit 203, the encoding unit 205 outputs a block header storing identification information (for example, “0”) indicating no replacement (non-pallet conversion). When a control signal indicating that the replacement process is valid is received from the determination unit 203, a block header storing the identification information (for example, “1”) indicating the presence of the replacement color (pallet conversion) and the palette conversion table is generated and output. To do. The palette conversion table stored in the block header may be the number of colors Nc and each color component value. The reason is that it is promised that the palette number for each color is 0, 1,..., Nc-1. FIG. 10 shows the structure of the encoded data file generated by the encoding unit 205 in the embodiment. The encoded data of one block is of any of the reference numbers 301 and 302 shown in the figure.

  Thereafter, the encoding unit 205 performs lossless encoding on each pixel data (color image data) in the raster scan order output from the replacement unit 204 according to the JPEG-LS encoding method, and generates encoded data. Then, the encoding unit 205 outputs the generated encoded data as data subsequent to the block header shown above.

  FIG. 6 is a detailed block diagram of the encoding unit 205 in the embodiment. As illustrated, the encoding unit 205 includes a context generation unit 251, a prediction assistant generation unit 252, a prediction error encoding unit 253, a run length conversion unit 254, a run length encoding unit 255, and a selector 256.

  As described above, the encoding unit 205 performs lossless encoding on the input image data according to the JPEG-LS encoding method. The JPEG-LS calculates the context value from the colors of the already encoded neighboring pixels Xa, Xb, Xc, and Xd in the vicinity of the target pixel X as shown in FIG. Then, according to the calculated context value, whether to perform predictive encoding or run-length encoding is determined. When the target block is on the first line of the block, or when the target pixel position is the left end or right end position of the line, all or part of the neighboring pixels Xa, Xb, Xc, Xd are positions outside the block. Become. Assume that the encoding process is performed assuming that the component values of the pixels outside the block are “0”. Here, the component value of the pixel outside the block is set to “0”. However, it may be the same as the decoding process, and thus may not necessarily be “0”.

  Now, as shown in FIG. 6, the context generation unit 251 internally includes two lines of memory in order to refer to pixel data of pixels near the pixel of interest X. Then, the context generation unit 251 reads neighboring pixels Xa, Xb, Xc, and Xd of the target pixel X from the internal memory, and calculates a context value at the target pixel X. Then, based on the context value, the context generation unit 251 determines whether the target pixel X should be encoded to be run-length or prediction error encoded, and the input target pixel X is converted into the prediction error calculation unit 252 and the run-length conversion. The data is output to one of the units 254.

  In the embodiment, for simplicity, whether or not the neighboring pixel Xa = Xb = Xc = Xd is satisfied, that is, the number of colors represented by the neighboring pixel {Xa, Xb, Xc, Xd} is “1”. Information indicating whether or not is a context value.

  When the number of colors combined by the neighboring pixels {Xa, Xb, Xc, Xd} is other than “1” (2 or more), the context generation unit 251 predictively encodes the target pixel X. The pixel component value is output to the prediction error calculation unit 252.

  Further, when the number of colors combined by the neighboring pixels {Xa, Xb, Xc, Xd} is “1”, the probability that the pixel of interest X is the same color as the immediately preceding pixel Xa is very high. Accordingly, the context generation unit 251 outputs each pixel component value of the target pixel X to the run length conversion unit 254 in order to start run-length encoding from the target pixel X.

  Once the target pixel X is output to the run length conversion unit 254, as long as the same pixel is input after the target pixel X, the context value is not calculated, and the input pixel data is run length converted. Output to the unit 254.

  In addition, during the run-length encoding process, when the pixel of interest X becomes the same color as the immediately preceding pixel Xa, or when the pixel of interest X reaches the right end of the block, the context generation unit 251 determines that the run has terminated. judge. Then, a signal indicating the end of the run is notified to the run length conversion unit 254, and the context value of the pixel of interest X is also calculated.

  The prediction error calculation unit 252 calculates the predicted value of each component of the target pixel from the neighboring pixels {Xa, Xb, Xc} of the target pixel X, and calculates the predicted value of each component and each component value of the target pixel. The difference is calculated and output to the prediction error encoding unit 253. In the embodiment, for simplicity, each component value in the neighborhood Xa is used as a predicted value of each component value of the pixel of interest. The prediction error encoding unit 253 generates a code word using a technique such as Huffman encoding or arithmetic encoding for the prediction error of each component of the pixel of interest, and outputs the code word to the selector 205.

  When the run length conversion unit 254 receives a signal indicating that the number of colors represented by the neighboring pixels {Xa, Xb, Xc, Xd} is “1” from the context generation unit 251, the run length conversion unit 254 starts counting the run. To do. Then, the run length conversion unit 254 indicates that the target pixel X has reached the right end of the block from the context generation unit 251 while the run is being counted. If a signal is received, the run counting is terminated. Then, the value counted so far is output to the run-length encoding unit 255. The run length encoding unit 255 generates a code word corresponding to the counted value input from the run length conversion unit 254 and outputs the code word to the selector 256.

  The selector 256 receives a signal indicating whether the run length encoding or the prediction error encoding is being performed on the target pixel by the context generation unit 251. Therefore, the selector 256 selects and outputs the encoded data from either the prediction error encoding unit 253 or the run length encoding unit 255 according to the received signal. Note that the prediction error encoding unit 253 outputs encoded data in units of one pixel, whereas the run length encoding unit 255 outputs encoded data only when one run ends in a run. . Therefore, the selector 256 controls input and output according to each timing.

  The configuration and processing related to encoding in the embodiment have been described above. Here, the reason why the palette is formed when the number Nc of colors appearing in the block is equal to or smaller than the threshold Th determined by the size of the block will be described below.

  The number Nc of colors appearing in one block being equal to or less than the threshold Th means that there are relatively few colors appearing in that block. When the number of colors is small, the probability that the difference between the colors is large tends to be high. In particular, in the case of CG (computer graphics), the number of colors in a block is relatively small, and this tendency is strong.

  In order to simplify the description, a monochrome image (lightness gradation image) will be described as an example. FIG. 8A shows an example in which there are four types of pixel values {C1, C2, C3, C4) appearing in the block, and the respective frequencies are in the order of C1> C2> C0> C4. Yes.

  When predictive error encoding is performed, a difference between a pixel value of interest (in this case, brightness) and an encoded pixel value (predicted value) in the vicinity thereof is calculated, and a code word corresponding to the difference is generated. In general, since the probability that the difference is small is high, a short codeword is generated when the difference is small, and a long codeword is generated as the difference value increases.

  However, in the state shown in FIG. 8A, the difference value is “0” if the pixel values are the same, but the difference value between the different pixel values is too large to be ignored. If the distance between the pixel values C0 and C1 is “50”, codewords with a difference of 1 to 49 are unused, and high coding efficiency cannot be expected.

  Therefore, a sequential palette number starting with 0 is assigned to the value of each pixel that actually appears. As a result, the histogram of FIG. 8A can be converted into the histogram of FIG. If this palette number is regarded as a pixel value, the colors C0, C1, C2, and C3 are replaced with palette numbers 0, 1, 2, and 3, respectively. Therefore, if this palette number is used, the absolute value of the prediction error is promised to be “3” at the maximum, and high coding efficiency can be expected.

  Furthermore, in the present embodiment, in the case of CG, there is a low probability that pixels of pixel values C0, C1, C2, and C3 are randomly arranged, and an area having pixel values of C0 → C1 → C2 → C3 (or the reverse) It is confirmed that there is a high probability that In other words, when calculating the prediction error, if the pixel value of the pixel of interest is C0, the probability that the predicted value is C0 is the highest, and then the order is C1, C2, and C3. When the pixel of interest is C1, the probability that the predicted value is C1 is the highest, followed by {C0, C2}, and then C3.

  To summarize the above, if the palette numbers are assigned in the order of appearance frequency, the probability that the prediction error can be reduced can be further increased. Therefore, in the present embodiment, when a histogram as shown in FIG. 8A is created, color data is rearranged using the frequency as a key instead of FIG. 8B, and the histogram shown in FIG. Create Then, a palette number is assigned to the rearranged pixel value. The determination unit 203 rearranges colors in descending order of appearance in the block of interest, and pairs the rearranged color data (RGB values) with palette numbers 0, 1, 2,..., Nc−1. The reason why the palette conversion table as shown in FIG. 5 is created is as described above. Note that FIG. 8C shows an example of rearranging in descending order using the frequency as a key, but it may be in ascending order.

  FIG. 11 shows details of the process in step S14 in FIG. 3 for performing the above process.

  First, in step S21, the control unit 150 determines the block size. In this process, if the block size is specified in the print data by the PDL data analysis process in step S11 (FIG. 3), the specified size is followed. If the block size is not specified in the print data, the size set by the operation unit 151 is followed. In step S21, a process for determining a threshold value for determining the number of colors is also performed according to the block size.

  Thereafter, in step S <b> 22, the blocking unit 200 of the image encoding processing unit 103 reads block data of the determined size from the buffer memory 101 and stores it in the buffer memory 201. In step S223, the histogram creation unit 202 scans the blocks stored in the buffer memory 201, creates a histogram, and outputs information about the created histogram to the determination unit 203.

  In step S24, the determination unit 203 calculates the number of appearing colors based on the input histogram information. In step S25, the determination unit 203 determines whether the calculated number of colors is equal to or less than the threshold set by the control unit 150. When determining that the calculated number of colors is greater than the threshold, the determination unit 203 outputs a control signal indicating that the calculated number of colors is greater than the threshold to the replacement unit 204 and the encoding unit 205. As a result, the encoding unit 205 generates and outputs a block header having identification information indicating no palette conversion in step S26. Further, the replacement unit 204 outputs the gradation image data in the buffer 201 to the encoding unit 205 as it is. Accordingly, in step S30, the encoding unit 205 encodes the input gradation image data as it is, and outputs the generated encoded data.

  On the other hand, when the determination unit 203 determines in step S25 that the number of colors is equal to or less than the threshold value, the process proceeds to step S27. In step S27, the determination unit 203 rearranges the colors of the histogram in order of frequency, and assigns palette numbers to the rearranged colors to create a palette conversion table. Then, the determination unit 203 outputs a signal indicating that the calculated number of colors is equal to or less than the threshold and information on the generated palette conversion table to the replacement unit 204 and the encoding unit 205.

When these pieces of information are input, the encoding unit 205 generates and outputs block information including identification information indicating that palette conversion has been performed, the number of colors, and data of each color component (step S28).

  The replacement unit 204 replaces the gradation image data in the buffer 201 with the palette number, and outputs the replaced data to the encoding unit 205 (step S29). Accordingly, the encoding unit 205 encodes the input palletized data and outputs the generated encoded data in step S30.

  When the process of step S30 is completed, the process proceeds to step S31, and the control unit 150 determines whether or not the encoding of all blocks has been completed. If not, the process after step S22 is performed to encode the next block. If it is determined that all blocks have been encoded, the encoding process for the page of interest ends.

  The encoding process according to this embodiment has been described above. Next, the decoding process will be described. Since the attribute flag decoding unit 107 need not be described, the image decoding processing unit 108 will be described below.

  FIG. 9 is a block diagram of the image decoding processing unit 108. The image decoding processing unit includes a header analysis unit 270, a context generation unit 271, a prediction error decoding unit 272, a run length decoding unit 273, a selector 274, a replacement unit 275, and a memory 276.

  The memory 276 has a capacity for storing at least two lines of decoding results of pixels near the target pixel X in one block. At the start of decoding for one block, the memory 276 is cleared to zero. This is because each component value of the pixel value outside the block is set to “0”.

  The header analysis unit 270 analyzes the identification information of the block header of the encoded block data, determines whether or not the block of interest has been encoded after replacement with the palette number, and if it has been converted to the palette number A process of extracting the number of colors and pixel values is performed. If it is determined that the data is the encoded data of the pallet number, the control signal indicating the encoded data of the pallet number is output to the replacement unit 275. Also, the pixel value stored in the block header and the palette number (how many pixel values exist depending on the number of colors can be easily found) are paired and output to the replacement unit 275. When it is determined that the target block is encoded as a gradation image, a control signal indicating that the gradation image is encoded is output to the replacement unit 275. Thereafter, the header analysis unit 270 outputs the encoded data subsequent to the block header to the context generation unit 271.

  When decoding the pixel of interest X, the context generator 271 refers to the memory 270 and calculates a context value from the neighboring pixels Xa, Xb, Xc, and Xd. Then, based on the calculation result, the context generation unit 271 determines whether the pixel of interest X is run-length encoded or prediction error encoded. In the embodiment, it is determined that predictive coding is performed when the number of colors indicated by four neighboring pixels is 2 or more, and run-length coding is performed when the number of colors is 1. When it is determined that the pixel of interest has been encoded with the prediction error (when the neighboring pixels have two or more colors), the context generation unit 271 outputs the encoded data to the prediction error decoding unit 272. When it is determined that the target pixel is run-length encoded (when the neighboring pixel has one color), the context generation unit 271 outputs the encoded data to the run-length decoding unit 273. The context generation unit 271 also outputs the determination result of the encoding type to the selector 275 as a control signal.

  The prediction error decoding unit 272 decodes the encoded data into a prediction error. Then, the prediction error decoding unit 272 refers to the memory 210, calculates a prediction value of the pixel of interest X, and adds a decoded prediction error to the prediction value, thereby decoding one component value of the pixel of interest X. And output to the selector 274. Then, this process is similarly performed for the remaining two components, thereby completing the decoding for one pixel.

  The run length decoding unit 273 decodes the run and outputs the run to the selector 275.

  The selector 275 outputs the decoding result from the prediction error decoding unit 272 to the replacement unit 275 and stores it in the memory 276 when a control signal indicating that the pixel of interest is subjected to prediction error encoding is input. When the selector 275 receives a control signal indicating that the target pixel is run-length encoded, the selector 275 sets the value of the pixel immediately before the run input from the run-length decoding unit 273 to the number indicated by the run. The data is output to the replacement unit 275 and the memory 276 by the amount.

  The replacement unit 275 outputs the pixel value from the selector 274 to the buffer memory 109 as it is when the control signal indicating that the block of interest is encoded as gradation image data is input from the header analysis unit 270. To do.

  On the other hand, it is assumed that a control signal indicating that the block of interest is encoded after being converted into a palette number, a palette number, and a pixel value are input from the header analysis unit 270. In this case, the replacement unit 275 pays attention only to the R component of the pixel value from the selector 274, assumes that the value of the R component is a palette number, replaces the corresponding pixel value, and outputs the result.

  As described above, according to the present embodiment, when an image is encoded in units of blocks, whether or not the image data is converted into a palette number and encoded according to the number of colors that appear in the block. By switching between, lossless encoding can be performed at a high compression rate.

<Second Embodiment>
In the above embodiment, the threshold value for determining whether or not to palletize is determined based on the block size. The images that are palletized to increase the encoding efficiency are character line drawings and computer graphics. Therefore, in addition to the block size, the attribute flag of each pixel in the block may be referred to determine whether to make a palette.

One attribute flag is generated for one pixel. Therefore, when the block size is m × n, the number of attribute flags is m × n. Therefore, when the number of character line drawing attribute flags + CG attribute flags (the number of pixels having non-natural image attributes) is M, the ratio Mr is calculated by the following equation.
Mr = 100 × M / (n × m)
If Mr is equal to or greater than the threshold value Th1 set in advance, the block of interest is determined to be palletized.

  FIG. 12 is a block configuration diagram of the image encoding processing unit 103 in the second embodiment. The difference from FIG. 4 is that the blocking unit 200 inputs image data of a size determined by a control signal from the control unit 150 and attribute flag data corresponding to the size from the buffer memory 101 and stores them in the buffer memory 201. It is a point to do. Moreover, the point which provided the ratio calculation part 206 in the structure of FIG. 4 and the point by which the processing content of the determination part 203 was changed are also different.

  The ratio calculation unit 206 refers to the attribute flag data for one block stored in the buffer memory 201 and counts the number M of pixels having a character / line drawing or computer graphics attribute. Then, the ratio Mr of M with respect to the block size (m × n) specified by the control signal from the control unit 150 is calculated, and the result is output to the determination unit 203.

  The determination unit 203 determines the threshold Th1 according to the block size. Then, it is determined whether or not the calculated ratio is equal to or greater than the threshold, that is, whether or not Mr ≧ Th1 is satisfied, and the determination result is output to the replacement unit 204 and the encoding unit 205. In particular, when Mr ≧ Th1, in order to palette-convert each pixel data of the block of interest, a histogram is created and a palette conversion table is created. Then, the generated palette conversion table is output to the replacement unit 204 and the encoding unit 205. If the calculated ratio is less than the threshold value, palette conversion is not performed.

  The subsequent processing is the same as in the first embodiment. There is no particular change in the data structure of the encoded data. The decoding process may be the same as in the first embodiment.

  As described above, according to the second embodiment, it is possible to achieve the same operational effects as those of the first embodiment.

<Third Embodiment>
In the first and second embodiments, when the replacement unit 204 executes palette conversion, the bus for transferring the information of the palette number uses one color component (R component in the embodiment) and other components (embodiments). In this example, the G and B component buses are encoded as “0”. When the palette conversion is performed, the G and B components are 0, so the encoded data of the G and B components are efficiently subjected to the length-length encoding. However, the components of G and B are still in line units. The encoded data is still generated. Therefore, when palette conversion is performed, only one component is used, so only the encoded data of the palette number may be output and the encoded data of the other two components may not be output. In this way, the compression rate can be further increased.

  In the above-described embodiment, an example applied to a printing apparatus has been described. Since this is a printing apparatus, a configuration for performing an encoding process and a configuration for performing a decoding process exist in one apparatus. However, each may be applied to a single device. For example, the present invention can also be applied to a case where an image is encoded by an information processing apparatus such as a personal computer, transferred to another apparatus for information processing, and decoded there. Further, in the case of an information processing apparatus such as a personal computer, naturally, the portions corresponding to the above-described processing units are made to function by computer program functions and subroutines, and the respective buffers are secured and used in the RAM. Therefore, the present invention also includes computer programs. Further, the computer program is usually stored in a computer-readable storage medium such as a CD-ROM, and can be executed by setting it in a reading unit (CD-ROM drive or the like) of the apparatus and copying or installing it in the system. Become. Therefore, it is obvious that the present invention also includes such a computer-readable storage medium.

  In the embodiment, an example in which JPEG-LS is used for image data encoding processing has been described. However, since only prediction error encoding may be used, JPEG-LS is not necessarily required.

It is a block diagram of a printer control unit in the embodiment. 1 is a block configuration diagram of a printing system in an embodiment. It is a flowchart which shows the main routine of the image coding process in embodiment. It is a block block diagram of the image coding process part of 1st Embodiment. It is a figure which shows the example of the palette conversion table which a determination part produces | generates. It is a block block diagram of the encoding part of FIG. It is a figure which shows the relative positional relationship of the pixel of interest and its vicinity pixel. It is a figure for demonstrating the conversion process from a histogram to a palette. It is a block block diagram of the image decoding process part in embodiment. It is a figure which shows the structure of the file of the encoding image data which the image encoding process part of embodiment produces | generates. It is a flowchart which shows the detail of step S14 of FIG. It is a block block diagram of the image coding process part of 2nd Embodiment.

Claims (13)

An image encoding device that encodes image data and generates encoded data,
Setting means for setting the block size of the encoding unit;
Determining means for determining a threshold according to the set block size;
Input means for inputting image data in units of blocks of a set size;
A histogram creation means for creating a histogram of the color of the input block of interest;
Comparing means for calculating the number of colors appearing in the block of interest from the created histogram, comparing the threshold value determined by the determining means with the number of colors, and determining whether the number of colors is equal to or less than the threshold value; ,
A conversion means for assigning a palette number to each color appearing in the block of interest and converting each pixel data in the block of interest into a palette number when the comparison means determines that the number of colors is equal to or less than the threshold;
When the comparison unit determines that the number of colors is equal to or less than the threshold value, the palette number converted by the conversion unit is losslessly encoded as pixel data, and the code obtained by encoding together with information indicating palette conversion. Output data,
If the comparison means determines that the number of colors is greater than the threshold value, each pixel data constituting the block of interest is losslessly encoded and the encoded data obtained by encoding is output together with information indicating non-palette conversion. An image encoding apparatus comprising: an encoding unit configured to The encoding means includes
When the comparison means determines that the number of colors is equal to or less than the threshold, the information indicating that the palette conversion has been performed, the number of colors that appear in the block of interest, and the correspondence between each palette number and pixel data The image encoding apparatus according to claim 1, wherein a table indicating the image data is stored in a block header, and the encoded data is stored subsequent to the block header. Furthermore, it comprises sorting means for sorting the histogram created by the histogram creating means in order of frequency,
3. The image encoding apparatus according to claim 1, wherein the conversion unit assigns a palette number to each color in accordance with an order after rearrangement by the rearrangement unit and performs palette conversion. Furthermore, it comprises generation means for interpreting print data described in a page description language and generating image data for one page,
4. The image encoding according to claim 1, wherein the input unit inputs the block-unit image data from one page of image data generated by the generation unit. 5. apparatus. A method of controlling an image encoding device that encodes image data and generates encoded data,
A setting step for setting the block size of the encoding unit;
A determination step of determining a threshold according to the set size of the block;
An input process for inputting image data in units of a set size block;
A histogram creation step for creating a histogram of the color of the input block of interest;
A comparison step of calculating the number of colors appearing in the target block from the created histogram, comparing the threshold value determined in the determination step with the number of colors, and determining whether the number of colors is equal to or less than the threshold value; ,
A conversion step of assigning a palette number to each color appearing in the block of interest and converting each pixel data in the block of interest into a palette number when it is determined in the comparison step that the number of colors is equal to or less than the threshold value;
If the number of colors is determined to be equal to or less than the threshold value in the comparison step, the palette number converted in the conversion step is losslessly encoded as pixel data, and the code obtained by encoding together with information indicating that the palette conversion has been performed Output data,
When it is determined in the comparison step that the number of colors is larger than the threshold value, each pixel data constituting the block of interest is losslessly encoded and the encoded data obtained by encoding is output together with information indicating non-palette conversion An image encoding device control method comprising: an encoding step for:   A computer program that causes a computer to function as the image encoding device according to any one of claims 1 to 4 when the computer reads and executes the computer program. An image encoding device that encodes image data and generates encoded data,
Setting means for setting the block size of the encoding unit;
Determining means for determining a threshold according to the set size of the block;
Input means for inputting image data and attribute data indicating whether each pixel constituting the image data belongs to a natural image or a non-natural image in units of blocks of a set size;
A comparison means for calculating a ratio of the number of pixels having an attribute indicating the non-natural image to the number of pixels in the block, and comparing the ratio with the threshold;
If the comparison means indicates that the ratio is greater than or equal to the threshold, a histogram creation means for creating a histogram of the color of the input block of interest;
Conversion means for assigning a palette number to each color appearing in the block of interest from the created histogram, and converting each pixel data in the block of interest into a palette number;
When the comparison means determines that the ratio is equal to or greater than the threshold value, the palette number converted by the conversion means is losslessly encoded as pixel data, and encoded data obtained by encoding together with information indicating that the palette conversion has been performed. Output
When the comparison means determines that the ratio is less than the threshold, each pixel data constituting the block of interest is losslessly encoded and outputs encoded data obtained by encoding together with information indicating non-palette conversion An image encoding apparatus comprising: means. The encoding means includes
When the comparison means determines that the ratio is equal to or greater than the threshold, the information indicating that the palette conversion has been performed, the number of colors that appear in the block of interest, and the correspondence between each palette number and pixel data are indicated. 8. The image encoding apparatus according to claim 7, wherein a table is stored in a block header, and the encoded data is stored subsequent to the block header. Furthermore, it has sorting means for sorting the histogram created by the histogram creating means in order of frequency,
The image encoding apparatus according to claim 7 or 8, wherein the converting unit performs palette conversion by assigning palette numbers in the order after rearrangement by the rearranging unit. Furthermore, the printing data described in the page description language is interpreted, generating image data for one page, and generating means for generating attribute data of each pixel when generating,
The said input means inputs the said block single image data and attribute data from the image data and attribute data for one page produced | generated by the said production | generation means, The any one of Claim 7 thru | or 9 characterized by the above-mentioned. The image encoding device described in 1. A control method for an image encoding apparatus that encodes image data and generates encoded data, and a setting step for setting a size of a block of an encoding unit;
A determination step of determining a threshold according to the set size of the block;
An input step for inputting image data and attribute data indicating whether each pixel constituting the image data belongs to a natural image or a non-natural image in units of blocks of a set size;
A comparison step of calculating a ratio of the number of pixels having an attribute indicating the non-natural image to the number of pixels in the block, and comparing the ratio with the threshold;
If the comparison step indicates that the ratio is equal to or greater than the threshold, a histogram creation step for creating a histogram of the color of the input block of interest;
A conversion step of assigning a palette number to each color appearing in the block of interest from the created histogram, and converting each pixel data in the block of interest into a palette number;
When the comparison step determines that the ratio is equal to or greater than the threshold value, the palette number converted in the conversion step is losslessly encoded as pixel data, and encoded data obtained by encoding together with information indicating palette conversion. Output
When the comparison step determines that the ratio is less than the threshold, each pixel data constituting the block of interest is losslessly encoded, and the encoded data obtained by encoding together with information indicating non-palette conversion is output. A control method for an image encoding device.   A computer program that causes a computer to function as the image encoding device according to any one of claims 7 to 10 when the computer reads and executes the computer program.   A computer-readable storage medium storing the computer program according to claim 6 or 12.

Images