JP5039425B2 - Image coding apparatus and control method thereof - Google Patents

Description

  The present invention relates to an image encoding technique.

  Conventionally, as a method of encoding an image, there has been a method such as JPEG encoding using orthogonal transformation. Here, JPEG encoding will be briefly described. First, an image is divided into 8 × 8 pixel blocks, and DCT (discrete cosine transform) is performed on each block. Subsequently, the 8 × 8 DCT coefficient values obtained by DCT are quantized using a quantization table. Since the quantization table usually increases the quantization step width for higher frequency components, the value after quantization becomes smaller for higher frequency components, and the frequency of occurrence of 0 increases.

  Subsequently, the quantized value is encoded, but the encoding process is different between the DC component and the AC component. As for the DC component, since the correlation between the blocks is strong, the differential prediction coding is performed, and then the Huffman coding is performed. Differential prediction encoding is a method for encoding a difference from the immediately preceding value. Huffman coding is a coding method that shortens the overall code length by giving short codes to values with high occurrence frequency and long codes to values with low occurrence frequency. For AC components, Huffman coding is performed after run-length coding. Run-length coding is a coding method in which the coding rate increases as the same value continues for a long time. Note that by performing zigzag scanning on the quantized values, the encoding efficiency is improved by making the zero value of the high frequency component continuous.

  JPEG encoding is a good compression method for natural images. On the other hand, however, JPEG encoding tends to generate mosquito noise due to loss of high-frequency components of images for images including characters and line drawings. Another problem is that block noise is likely to occur due to the loss of quantization error of DC components and loss of high frequency components.

  On the other hand, in Patent Document 1, an edge enhancer that enhances the edge component of input image data enhances the edge component of input image data in accordance with the contents of the quantization table used by the quantization means. Thus, it is disclosed that noise is reduced.

Further, in Patent Document 2, the determination that an image is an important region and the other is provided, and for each block subjected to orthogonal transform and quantization processing, the quantization coefficient of a block that does not belong to the important region is replaced with 0. It is disclosed that a high image quality is partially maintained.
JP 09-172636 A Japanese Patent Application Laid-Open No. 06-164941

  However, in encoding so far, the position of the upper left corner in a certain block is set as the starting point, and raster scanning or zigzag scanning is performed from the starting point position to perform encoding processing. To increase the coding efficiency, the probability that the same value continues during the scan is increased. However, in the conventional encoding technique, the scan end position is the position of the lower right corner of the block in either the raster scan or the zigzag scan. In other words, if there is data that is different from the previous data in the lower right corner, it is necessary to generate encoded data of all data, and there is still room for improvement in terms of encoding efficiency. It was.

  The present invention has been made in view of this point, and intends to provide a technique for further improving the coding efficiency.

In order to solve this problem, for example, an image encoding device of the present invention has the following configuration. That is,
Input means for inputting a plurality of pixel data representing an image block;
A plurality of encoding processing means for generating candidate encoded data for the target image block by scanning and encoding the plurality of pixel data in the target image block input by the input unit;
Among the plurality of candidate coded data generated by said plurality of encoding means, selection means for selecting a candidate coded data is the minimum amount of data,
Identification information indicating which of the plurality of candidate encoded data is selected, and output means for outputting the selected candidate encoded data as encoded data of the target image block ,
All of the encoding processing means for generating the candidate encoded data scan and encode according to scan routes having different scan start positions.

  According to the present invention, a plurality of encoding processing means rearrange pixel data in a block according to different scan start positions and scan routes, respectively, and generate encoded data. And the code data which is the minimum data amount in it is output. As a result, the amount of code data can be greatly reduced.

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

<First Embodiment>
FIG. 6 shows a block configuration diagram of the image encoding device according to the first embodiment.

  The block division unit 600 inputs image data (hereinafter referred to as image block data) composed of m × n pixels from the image data to be encoded. The input order of the image block data is a raster scan order in block units. In order to simplify the description, the embodiment will be described assuming that the image block has a size of 8 × 8 pixels. The source of the image data to be encoded is an image scanner, but it may be a storage medium that stores the image data, or a rendering unit that generates image data based on print data. The type is irrelevant.

  The apparatus includes a plurality of encoding processing units that input image block data and generate encoded data. Reference numerals 610, 620, 630, and 640 shown in the figure are the encoding processing units. Here, four examples of the encoding processing unit are shown, but the present invention is not limited by the number.

  Each of the image encoding processing units 610, 620, 630, and 640 encodes the image block data supplied from the block dividing unit 600, generates encoded data, and outputs the generated encoded data to the selector unit 650. .

  The selector unit 650 selects one of the encoded data output from the image encoding processing units 610, 620, 630, and 640 having the smallest amount of encoded data (minimum number of bits). Then, selector section 650 generates an identification bit indicating which encoded data has been selected, outputs the identification bit first, and then outputs the selected encoded data. In the case of the embodiment, since there are four image encoding processing units, it is sufficient that there are two identification bits.

  The outline of the processing of the image coding apparatus in the present embodiment is as described above. Hereinafter, the image coding processing unit 610 will be described in more detail. As will become clear from the following description, the image encoding processing units 610, 620, 630, and 640 are the same in hardware. However, different information is stored in each of the scan start position information storage unit and the scan route information storage unit. Further, the scan start position information storage unit and the scan route information storage unit are composed of rewritable memories (registers). The reason for this is that it is not necessary to separately design the four image encoding processing units, and the development cost can be reduced accordingly. In the description of the embodiment, it is assumed that the same information is stored in the prediction error allowable value storage units in the four image encoding processing units.

  The scan conversion unit 611 has an internal buffer for storing one image block data (8 × 8 pixel data). When the scan conversion unit 611 stores the image block data of 8 × 8 pixels from the block division unit 600 in the internal buffer, the scan conversion unit 611 starts the pixel array conversion process. First, the scan conversion unit 611 reads the scan start position information (data in the x-coordinate and y-coordinate formats) stored in the scan start position information storage unit 614, reads the pixel data of the internal buffer indicated by the position, and performs prediction The data is output to the encoding unit 612. Thereafter, the scan conversion unit 611 sequentially reads the relative position information stored in the scan route information storage unit 615, and sequentially updates the address for accessing the internal buffer. Each time the address is updated, the internal buffer is accessed to read out the pixel data and output it to the predictive coding unit 612. In this way, the scan conversion unit 611 refers to the input 8 × 8 pixel data with reference to the scan start position information storage unit 614 and the scan route information storage unit 615, and the 64 pieces of the one-dimensional array in which the pixel data is rearranged. Are output to the predictive encoding unit 612. The predictive encoding unit 612 encodes the pixel data after the rearrangement change. Note that the reason for storing the relative address in the scan route information storage unit 615 is that if the scan order always passes through adjacent pixels, only -1, 0, +1 may be stored for both the X and Y addresses. This is because the storage capacity can be reduced. Since it is only necessary to specify any of the three X and Y addresses, it can be said that two bits are sufficient for the relative addresses of the X and Y coordinates.

  FIG. 7A shows a scan route by the scan conversion unit 611 in the image encoding processing unit 610. In the illustrated case, the scan start position is the upper left corner of the image block. Then, when it reaches the right end from the start position in the right direction, it repeatedly scans one lower line along the route from the right end to the left end.

  FIGS. 7B to 7D show scan start positions and scan routes in the respective scan conversion units in the encoding processing units 620 to 640. As shown in the figure, it can be understood that the scan start positions of the scan conversion units in the four encoding processing units and their scan routes are different from each other.

  Next, the process of the predictive coding unit 612 will be described. The predictive encoding unit 612 generates encoded data of the 64 pixel data input from the scan conversion unit 611 in the input order and outputs the encoded data.

  The encoding algorithm of the predictive encoding unit 612 is as follows.

  The predictive encoding unit 612 has a counter inside, and clears the counter to zero when starting encoding of the image data (64 pixel data) of the block of interest. For the first pixel data, the pixel data is output and the value of the pixel data is held as a reference value. When the second and subsequent pixel data are input, an absolute value of a difference from the reference value (hereinafter simply referred to as a difference value) is calculated, and the difference value is stored in the prediction error allowable value storage unit 616. It is determined whether the error is less than the allowable value. When the difference value is equal to or smaller than the prediction error allowable value, the pixel data of interest is not output and the counter is incremented by “1”. When it is determined that the difference value between the target pixel data and the reference value exceeds the prediction error allowable value, first, encoded data based on the value held in the counter is output. Then, the encoded data of the target pixel data is output, and the value of the target pixel data is updated as a new reference value. At this time, the counter is cleared to zero. Even if the target pixel is the final pixel position (the 63rd pixel when the first pixel is the 0th pixel), if it is determined that the difference value is equal to or smaller than the prediction error allowable value, the skipped number is used instead. A preset EOB (End Of Block) codeword is output.

  Hereinafter, an example of the process of the prediction encoding unit 612 will be described. Now, assume that the 8 × 8 pixel image block data to be encoded is shown in FIG. Further, it is assumed that the prediction error allowable value stored in the prediction error allowable value storage unit 616 is “3”.

Here, the predictive encoded data generated and output by the predictive encoding unit 612 is represented as (A, B). Here, the meanings of A and B are as follows.
A: “0” if the difference value is larger than the prediction error tolerance, “1” if the difference is less than the prediction error tolerance;
B: Input pixel data when the difference value is larger than the prediction error allowable value, and data counted until then if the difference value is less than the prediction error allowable value;
However, even if the last pixel in the block is input, if the difference value is smaller than the prediction error allowable value, “EOB” is set.

Now, the encoded data when the image block data in FIG. 7E is scanned by the scan route in FIG. 7A is as follows.
“(0,0) (1,53) (0,23) (1,3) (0,12) (0,0) EOB”
The variable length encoding unit 613 generates further binary encoded data from the encoded data, and outputs it to the selector unit 650. The variable length encoding unit 613 may use any encoding technique.

Also, the predictive encoding unit in the image encoding processing unit 620 generates encoded data by scanning with the scan route of FIG. The encoded data generated in this case is as follows.
“(0,2), (1,47), (0,20), (1,1), (0,2), (1,10), (0,12), (0,20), ( 1,1) ",
Also, the predictive coding unit in the image coding processing unit 630 scans with the scan route in FIG. 7C to generate coded data. The encoded data generated in this case is as follows.
“(0,21), (1,1), (0,12), (0,0), (1,10), (0,23), (1,1), (0,2), EOB "
Then, the predictive coding unit in the image coding processing unit 640 scans with the scan route in FIG. 7D to generate coded data. The encoded data generated in this case is as follows.
“(0,0), (1,4), (0,12), (0,20), (1,3), (0,2), EOB”

  It should be noted that when an EOB code word is output from the predictive encoding unit 612, the data indicated by the pixel data indicated by the code word immediately before the EOB continues to the last pixel. It is a point that can be determined to be. In other words, it can be said that the earlier the timing at which an EOB codeword is generated from the predictive encoding unit 612, the smaller the amount of encoded data generated by the variable length encoding unit 613. Further, when the prediction error allowable value stored in the prediction error allowable value storage unit 616 is set to “0”, the encoding processing unit 610 generates lossless encoded data. Therefore, the value stored in the prediction error allowable value storage unit 616 is advantageously configured by a writable memory so that it can be set according to a user instruction.

  By repeating the above procedure for all blocks, the encoding for the entire area of the image data to be encoded is completed. Note that the selector 650 selects one of the four encoded data generated from the block of interest that has the shortest code length, and the selected encoded data is generated by any encoding processing unit. Identification information (2 bits) indicating whether or not is added to the head and output. Also, at the start of the encoding process, a file containing information necessary for the decoding process, such as the size of the image to be encoded (number of pixels in the horizontal and vertical directions), the accuracy of each pixel (number of bits), the type of color space A header will be generated.

  Next, processing procedures of the scan conversion unit 611 and the predictive encoding unit 612 in the embodiment will be described with reference to the flowchart of FIG. FIG. 8 shows a processing procedure for one image block.

  First, in step S801, the scan conversion unit 611 converts the arrangement of pixels constituting the input image block data based on the information stored in the scan start position information storage unit 614 and the scan route information storage unit 615. . Thereafter, the process proceeds to step S802. Note that the processing after step S802 is the processing of the predictive encoding unit 612.

  Next, in step S802, the prediction encoding unit 612 clears the internal counter CNT to zero.

  In step S803, a difference value between the target pixel data input from the scan conversion unit 611 and the reference value is calculated, and it is determined whether the difference value is equal to or less than a prediction error allowable value. When the target pixel is the first pixel, it is determined that the difference value exceeds the prediction error allowable value, and the value “0” of the counter CNT is output.

  If it is determined that the difference value between the target pixel data and the reference value is equal to or less than the prediction error allowable value, the process proceeds to step S804, and it is determined whether the target pixel is the final pixel. If it is determined that the pixel is not the final pixel, the process proceeds to step S805, the counter CNT is incremented by “1”, the next pixel data is read, and the process of step S803 is performed. If it is determined in step S804 that the pixel-of-interest data is the last pixel, an EOB code word is output and the process is terminated.

  On the other hand, if it is determined in step S803 that the difference value between the target pixel data and the reference value is larger than the prediction error allowable value, the process proceeds to step S807. In step S807, it is determined whether or not the value of the counter CNT is “0”.

  If it is determined that the value of the counter CNT is “0”, the target pixel data is output in step S809.

  If it is determined in step S807 that the value of the counter CNT is other than “0”, the value of the counter CNT is output in step S808, and the target pixel data is output in subsequent step S809. .

  The process proceeds to step S810, and it is determined whether the target pixel is the last pixel. If it is determined that the target pixel is not the final pixel, the value of the counter CNT is set to “0”, and the process returns to step S803. If it is determined that the target pixel is the final pixel, the encoding process for one image block is terminated.

  As described above, according to this embodiment, the four image encoding processing units 610, 620, 630, and 640 generate encoded data according to different scan routes from different start positions in the image data block. . These image encoding processing units generate an EOB code word when the target pixel reaches the final pixel while the difference value (prediction error) of the target pixel data remains within the allowable range. Then, by selecting the shortest encoded data generated by the four encoding processing units, it is possible to generate encoded data with a high compression rate.

  In addition, what is necessary is just to follow the procedure contrary to the above, when decoding the said encoded data. That is, first, the first decoding unit decodes the encoded data following the first 2-bit identification information, and decodes even the encoded data generated by the predictive encoding unit 612. Then, the second decoding unit generates one-dimensional 64 pixel data. Thereafter, 64 pixel data may be arranged and output in accordance with the start position specified by the identification bit and the scan route.

  In the embodiment, the size of the image block is 8 × 8 pixels, but the present invention is not limited by this size, and is generally applicable to m × n pixels (m and n are integers of 2 or more). It is. Also, the number of encoding processing units is four and the scan routes are also shown in FIGS. 7A to 7D. However, the number of encoding processing units and the scan route are not limited to this. In particular, as described in the embodiment, it can be easily understood that the scan start position and the scan route can be freely set by changing each information.

<Second Embodiment>
A second embodiment will be described. FIG. 1 is a block configuration diagram of an image encoding device according to the second embodiment.

  In the figure, the preprocessing unit 101 inputs image data in units of blocks and performs a series of processes until it is separated into an extracted color part and a background part (non-extracted color part) in the input block. Also in the second embodiment, the block size is 8 × 8 pixels, but may be 16 × 16 pixels, 32 × 32 pixels, or the like.

  An example of a method for separating the extracted color portion and the background portion is as follows.

First, the average value of the density of pixel data in the block is calculated. Then, the pixel group is classified into a pixel group having a value larger than the average value (hereinafter referred to as a first pixel group) and a pixel group having a value equal to or less than the average value (hereinafter referred to as a second pixel group). Then, an average value of the first pixel group (first average value) and an average value of the second pixel group (second average value) are obtained, and the absolute value of the difference between them is set to a preset threshold value. It is determined whether or not it exceeds. If it exceeds, a control signal indicating that classification is possible is notified to the pack unit 105 . Then, the first pixel group is regarded as a pixel group constituting the extraction color portion. The second average value is used as replacement information described later.

  When the absolute value of the difference between the first average value and the second average value is equal to or smaller than a preset threshold value, the preprocessing unit 101 performs control indicating that separation into the extraction pixel unit and the background pixel unit is impossible. The signal is notified to the pack unit 105.

  Here, the data of the extracted color portion includes position information (position data) indicating the position of the pixel having the extracted color, and extracted color information. This position information is binary 1-bit data that is “1” at the pixel position belonging to the extracted color portion and “0” at the pixel position belonging to the background pixel portion (non-extracted color portion). That is, the position information can be said to be information for identifying whether each pixel in the block is an extracted color pixel or a non-extracted color pixel. In this embodiment, since the size of one block is 8 × 8 pixels, the position data is 8 × 8 = 64 bits. The order of bits is in raster scan order. This position data is encoded by the first encoding unit 102 in the subsequent stage. The extracted color information is encoded by the second encoding unit 103 at the subsequent stage.

  On the other hand, the background portion is 8 × 8 pixel data obtained by replacing the value of each pixel determined as the extracted color portion in the block with the replacement information (second average value) described above. Further, the replacement processing is not performed on the pixels determined as non-extracting portions in one block. As a result, the high-frequency component is not included in the image data (image data after replacement) indicated by the block indicated by the background portion, or the ratio of the high-frequency component is low. This background part is irreversibly encoded by a third encoding unit 104 (for example, configured by a DCT transform unit, a quantization unit, and entropy encoding).

  The encoding units of the first encoding unit 102, the second encoding unit 103, and the third encoding unit 104 can employ independent encoding algorithms.

  The pack unit 105 multiplexes and outputs the data encoded by each encoding unit. However, as described above, when the signal indicating that the extraction pixel unit and the background pixel unit cannot be separated is received from the preprocessing unit 101, the pack unit 105 outputs only the encoded data from the encoding unit 104. To do. In order to identify whether one block is composed of three types of encoded data or one type of encoded data, the pack unit 105 is placed at the head (block header) of one block of encoded data. Stores the identification bit.

Now, features of the second embodiment, in the first encoding unit 102 for encoding the position data shown above. The first encoding unit 102 will be described in detail below.

FIG. 2 is a block configuration diagram of the first encoding unit 102 in the second embodiment.

As shown in the figure, the internal configuration of the first encoding unit 102 is substantially the same as that of FIG. 6 shown in the first embodiment. That is, the first encoding unit 102 includes a plurality (four in the embodiment) of encoding processing units 210, 220, 230, and 240. Then, the selector unit 250 compares the data amounts of the encoded data from the four encoding processing units, and selects the encoded data having the minimum data amount. Then, the selector unit 250 outputs the selected encoded data following the identification information (ID information) for identifying which one is selected.

  Since the four encoding processing units are the same in hardware, only the encoding processing unit 210 will be described here.

  The scan conversion unit 201 receives the extracted pixel position data (8 × 8 = 64 bits) supplied from the preprocessing unit 101. The scan conversion unit 201 scans the position information based on the scan start position information stored in the scan start position information unit 213 and the scan route information stored in the scan route information storage unit 214 (reordering is performed). Output). That is, the position information is rearranged. The rearranged position information is supplied to the variable length encoder 212. The variable length encoding unit 212 performs variable length encoding on the input position information, and outputs the generated encoded data to the selector unit 250.

  The above is for the encoding processing unit 210, but the other encoding processing units 220, 230, and 240 are the same. The difference is the information set in the scan start position information section and the scan route information storage section by the scan conversion section in each encoding processing section.

  As described above, the selector unit 250 receives the variable length encoded data by the encoding processing units 210, 220, 230, and 240, compares the code amounts of the respective data, and determines the code with the smallest code amount. Select one data. Then, the identification information for identifying whether the selected encoded data is from the encoding processing unit 210, 220, 230, or 240 is set at the head, and subsequently the selected encoded data is output. .

  Here, as shown in FIG. 3A, the scan conversion unit 211 in the encoding processing unit 210 in the second embodiment sets the upper left corner position as the scan start position. Then, the scan conversion unit 211 scans in the right direction from the start position, and when it reaches the right end, it repeats scanning in the left direction from the right end of one line (horizontal zigzag scan). The scan conversion unit 211 reads position information according to the scan route (scan pattern) and outputs the position information to the variable length coding unit 212.

  Each of the scan conversion units in the other encoding processing units 220, 230, and 240 performs rearrangement of position information according to different scan start positions and different scan routes as shown in FIGS.

  As described above, since the position information of one pixel is binary, the variable-length encoding unit 212 encodes binary data, and uses, for example, run-length encoding. The variable-length encoding unit 212 generates encoded data in a format such as “run length of value“ 0 ”, run length of value“ 1 ”, run length of value“ 0 ”,. To do.

  However, if the same value continues until the final scan position, an EOB code word is added instead of the run-length encoded data and output. Here, EOB is an abbreviation for End Of Block.

  Now, when the position information of 8 × 8 bits is as shown in FIG. 3E, the encoded data generated by the variable length encoding unit 212 based on the scan route of FIG. 5, EOB ". Also, the encoded data in the case of the scan route in FIG. 3B is “48, 2, 11, 3”, and the encoded data in the case of the scan route in FIG. 3C is “3, 11, 2, EOB”. 3], the encoded data in the case of the scan route in FIG. 3D is “5, 5, EOB”.

  However, if the same value continues until the final scan position, an EOB code word is added instead of the run-length encoded data and output.

  The above procedure is repeated for all blocks.

  In the case of decoding, if the position information immediately before the discovery of the EOB code word is “1” and 30-bit position information has been decoded so far, the remaining 34 (= 64− 30) The bit may be regarded as a value of “0” and decoded.

  Next, processing of the encoding processing units 210, 220, 230, and 240 and the selector unit 250 will be described with reference to the flowchart of FIG.

  First, in step S401a, the scan conversion unit 211 of the encoding processing unit 210 performs scan conversion of the extracted pixel position information based on the respective information stored in the scan start position information storage unit 213 and the scan route information storage unit 214. (Sort position information). The scan conversion units of the other encoding processing units 220, 230, and 240 perform the same processing in steps S401b, 401c, and 401d.

  Next, in step S402a, the variable length coding unit 212 of the coding processing unit 210 performs variable length coding of the extracted pixel position information after the rearrangement. At this time, if the position information has not changed before reaching the last position information, an EOB codeword is set instead of the run-length encoded data of the position information. The variable length coding units of the other coding processing units 220, 230, and 240 perform the same processing in steps S402b, 402c, and 402d.

  When the encoding process for one block in the encoding processing units 210, 220, 230, and 240 is completed, the process proceeds to step S403. In step S403, the selector unit 250 compares the encoding amount (encoded data length) generated by each encoding processing unit, and selects the one having the smallest code amount. In step S404, the selected encoded data is output with the identification information indicating which encoded data has been selected as the head. In the case of the embodiment, since the number of encoding processing units is four, two bits are sufficient for this identification information.

  In step S405, it is determined whether there is a block that has not yet been encoded. If there is an unencoded block, the processes in steps S401a, 401b, 401c, and 401d are repeated. If the encoding process has been completed for all blocks, the process ends.

  As described above, according to the second embodiment, it is possible to efficiently encode position information (binary data) indicating an extracted color position when image data is encoded in units of blocks. .

<Third Embodiment>
In the first embodiment, the scan pattern shown in FIGS. 7A to 7D is described. In the second embodiment, the scan pattern shown in FIGS. 3A to 3D is used. This does not limit the present invention. In short, any start position and scan route can be set depending on the information set in the scan start position information storage unit and the scan route information storage unit.

  For example, for example, the scan start position and the scan route shown in FIGS. 5A to 5D may be adopted.

  FIG. 5A shows an example of zigzag scanning in the downward direction with the scan start position as the upper left corner. FIG. 5B shows an example in which the scan start position is the lower left corner and zigzag scanning is performed in the right direction. FIG. 6C shows an example of zigzag scanning in the upper left direction with the scan start position as the lower right corner. FIG. 4D shows an example of zigzag scanning in the lower left direction with the scan start position as the upper right corner. These settings can be realized by changing information in the scan start position information storage unit 213 and the scan route information storage unit 214 that stores relative coordinates from the scan start position.

  Consider a case where the scan patterns of FIGS. 5A to 5D are applied to the second embodiment. However, the variable-length encoding unit in each encoding processing unit is configured such that “0 run length, 1 run length, 0 run length, 1 run length,. The encoded data of the same run (same data) are output alternately. However, when the same data continues up to the final position information, an EOB code word is output instead of the run length code word.

  In this case, when the data to be encoded (position information) is FIG. 5E, the encoded data when encoded according to the scan pattern 501 of FIG. 5A is “40, 1, 13, 1, 1, 3, EOB ". Also, the encoded data when encoded according to the scan pattern 502 of FIG. 5B is “47, 3, 11, 1, 1, 1”. The encoded data when the scan pattern 503 of FIG. 5C is used is “2, 1, 3, EOB”, and when the scan pattern 504 of FIG. 6D is used, “15, 1, 12 2, 12, 2, EOB ". As can be seen from this result, it can be understood that the amount of encoded data when the scan pattern 503 in FIG. 5C is used can be sufficiently reduced as compared with other encoded data.

  Considering the above, several combinations of scan start position information and scan route information to be set in the four encoding processing units are prepared, and information indicating which one has been selected by the user is included in the file header of the encoded file. It is conceivable to store them.

For example, a scan start position and a scan route set in one encoding processing unit are expressed as (SS, SR). In the embodiment, since there are four encoding processing units, the scan start position and scan route information set in the four encoding processing units are {(SS1, SR1), (SS2, SR2), (SS3, SR3), (SS4, SR4)}. These four combinations are defined as P (i) using a variable i (i is 0, 1, 2,...).
P (i) = {(SS1 (i), SR1 (i)), (SS2 (i), SR2 (i)), (SS3 (i), SR3 (i)), (SS4 (i), SR4 ( i))}

  When encoding is performed, the variable i is changed in order from 0, and the encoding process is performed each time. Then, the variable i having the smallest code amount is determined, and the variable i is stored in the file header of the encoded data and output.

  First, the decoding device analyzes the file header and extracts the variable i, thereby obtaining four patterns {(SS1, SR1), (SS2, SR2), (SS3, SR3), (SS4, SR4)}. It should be decided.

  As described above, while using four encoding processing units, it is possible to perform encoding processing based on a scan start position and a scan route of more combinations.

  The above may also be reflected as it is in the first embodiment. As described above, according to the third embodiment, it is possible to further increase the coding efficiency as compared with the first and second embodiments.

  The embodiment according to the present invention has been described above. In the present embodiment, four examples of the number of encoding processing units have been shown, but the number is not limited to this number. In short, the encoding target data may be rearranged according to different scan start positions and different scan routes in a plurality of encoding processing units.

  Further, it is obvious that the functions shown in the above embodiments can be realized by a computer program executed by a general-purpose information processing apparatus such as a personal computer. If the OS (operating system) supports multitasking, one block of encoding processing may be performed in parallel. In addition, when the encoding processes cannot be performed in parallel, the encoding processes are performed in order for each block, and then the comparison process is performed.

  Also, 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 reader (CD-ROM drive or the like) and copying or installing it in the system. Become. Therefore, it is clear that such a computer readable storage medium is also within the scope of the present invention.

It is a block block diagram of the image coding apparatus in 2nd Embodiment. It is a figure which shows the structure of the 2nd encoding part in 2nd Embodiment. It is a figure which shows an example of the scan route of each encoding process part in 2nd Embodiment, and the data of an encoding target. It is a flowchart which shows the process sequence of the 2nd encoding part in 2nd Embodiment. It is a figure which shows an example of the scan route in another embodiment, and the encoding object data. It is a block block diagram of the image coding apparatus which concerns on 1st Embodiment. It is a figure which shows an example of the scan route of each encoding process part in 1st Embodiment, and the data of encoding object. It is a flowchart which shows the encoding process procedure of each encoding process part in 1st Embodiment.

Claims (7)

Input means for inputting a plurality of pixel data representing an image block;
A plurality of encoding processing means for generating candidate encoded data for the target image block by scanning and encoding the plurality of pixel data in the target image block input by the input unit;
Among the plurality of candidate coded data generated by said plurality of encoding means, selection means for selecting a candidate coded data is the minimum amount of data,
Identification information indicating which of the plurality of candidate encoded data is selected, and output means for outputting the selected candidate encoded data as encoded data of the target image block ,
An image coding apparatus characterized in that all of the coding processing means for generating the candidate coded data scan and code according to scan routes having different scan start positions. Each of the plurality of encoding processing means includes:
When the pixel value of the pixel at the scan start position is an initial reference value, and the difference between the pixel value of the target pixel obtained according to the scan route and the reference value is within a preset allowable range, Count the number of pixels from the pixel set as the reference value to the target pixel,
When the difference between the pixel value of the target pixel and the reference value exceeds the allowable range, the value counted so far is output as encoded data, and the pixel value of the target pixel is set as a new reference value And outputting encoded data of the pixel value of the pixel of interest,
When the difference between the pixel value of the pixel of interest and the reference value remains within the allowable range and the pixel of interest reaches the position of the final pixel, encoded data indicating an EOB (End Of Block) codeword is obtained. The image encoding device according to claim 1, wherein the image encoding device outputs the image. Each of the plurality of encoding processing means includes:
Rewritable scan start position storage means for storing information indicating the scan start position;
Rewriteable scan route storage means for storing a relative address indicating any one of {-1, 0, +1} in both the horizontal and vertical directions with respect to the position of the pixel scanned immediately before in order to indicate the scan route The image encoding device according to claim 1, wherein:   The image encoding apparatus according to claim 2, further comprising setting means for setting the allowable range. An input step of inputting a plurality of pixel data representing an image block;
A plurality of encoding processing steps for generating candidate encoded data for the target image block by scanning and encoding the plurality of pixel data in the target image block input in the input step;
A selection step of selecting candidate encoded data having a minimum data amount among the plurality of candidate encoded data generated in the plurality of encoding processing steps;
Identification information indicating which of the plurality of candidate encoded data is selected, and an output step of outputting the selected candidate encoded data as encoded data of the target image block ,
A method for controlling an image encoding device, characterized in that all encoding processing steps for generating candidate encoded data are scanned and encoded according to scan routes having different scan start positions.   A computer program that causes a computer to function as each unit included in the image encoding device according to any one of claims 1 to 4 when the computer reads and executes the computer program.   A computer-readable storage medium storing the computer program according to claim 6.

Images