JP2009111821A - Image encoding device, image decoding device, image data processing device, image encoding method, and image decoding method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an image encoding device capable of further reducing the amount of encoded compressed data generated when image data is compressed by BTC encoding.
An image data dividing unit that divides image data into unit blocks composed of a plurality of pixels, an average value calculating unit that calculates an average value of each element value in the unit block, and a degree of variation in distribution of each element value. A variation degree determining means 13 for determining the magnitude and generating a flag bit, a comparison result data generating means 14 for generating comparison result data indicating each size comparison result of each element value and the average value, each element value and each average value Deviation value calculation means 15 for calculating a deviation value based on the difference value and encoded data according to the value of the flag bit are generated by combining the average value and the flag bit when the variation degree is small. In the case of a larger value, there is provided encoded data generation means 16 for generating an average value, a flag bit, comparison result data, and a deviation value.
[Selection] Figure 1

Description

  The present invention relates to an image encoding apparatus and method for compressing image data, and an image decoding apparatus and method for decompressing compressed image data. Further, the present invention relates to an image using the image encoding apparatus and the image decoding apparatus. The present invention relates to a data processing apparatus.

  As a technique for dividing input image data into small pixel blocks and coding and compressing them, a BTC (Block Truncation Coding) coding method has been devised and applied to image display devices (for example, Patent Document 1 and Patent Document). 2).

  The procedure and principle of encoding processing and decoding processing according to the conventional BTC encoding method are shown in FIGS. In the conventional BTC encoding method, image data composed of a plurality of pixels is divided into small pixel blocks of N × N (normally N = 2 or 4), and processing is performed in units of the pixel blocks. In the following description, it is assumed that N = 2.

  As shown in FIG. 15A, first, the image data is divided into 2 × 2 pixel blocks (step # 1). Here, as shown in FIG. 15B, each pixel value is represented by Pij (i = 0, 1, j = 0, 1). An average value AVG of four pixel values of each pixel block (in the case where each pixel is composed of RGB component gradation data or luminance data, the pixel value indicates each RGB component or luminance gradation value), It calculates based on the following formula 1 (step # 2).

  Subsequently, the four pixel values Pij of each pixel block are compared with the average value AVG. When the pixel value Pij is larger than the average value AVG, it is encoded as “0”, and when it is smaller, it is encoded as “1”. 2 "and 2 columns of comparison result data Cij (i = 0, 1, j = 0, 1) are created (step # 3). Next, a deviation value SD indicating a deviation from the average value AVG of the four pixel values Pij of each pixel block is calculated based on the following equation (2) (step # 4).

  Then, as shown in FIG. 16, the average value AVG calculated in step # 2, the comparison result data Cij created in step # 3, and the deviation value SD calculated in step # 4 are synthesized, and each pixel block Encoded data is generated (step # 5) and stored in the image memory (step # 6).

  The processes from step # 1 to step # 6 are repeated for all pixel blocks.

  Next, a processing procedure for decoding the encoded data generated in the above manner and stored in the image memory into the data format of the original image data and displaying the image will be described with reference to FIG.

  First, the average value AVG, the comparison result data Cij, and the deviation value SD are read out from the encoded data of each pixel block stored in the image memory (step # 7). Next, as shown in FIG. 17B, the decoded pixel value of each pixel block is represented by P′ij (i = 0, 1, j = 0, 1), and each pixel value P′ij Is generated in the following manner using the average value AVG, the comparison result data Cij, and the deviation value SD (step # 8). When the comparison result data Cij is “0”, P′ij = AVG + SD, and when the comparison result data Cij is “1”, P′ij = AVG−SD. The decoded pixel value P′ij has the same data format and data length (8-bit data) as the pixel value Pij of the pixel block of the image data before encoding. The processes in steps # 7 and # 8 are repeated for all the pixel blocks, and the decoded image data of each pixel block is output to the image display device and displayed (step # 9).

  FIG. 18 shows the encoded data generated by the encoding process shown in FIG. 15A and the decoding process shown in FIG. 17A for specific numerical values of the pixel values Pij of one pixel block. An example of each pixel value P′ij of one pixel block to be performed is shown. In FIG. 18, the pixel values Pij of the pixel block of 2 × 2 pixels assuming 8-bit data (256 gradations) are P00 = 120, P01 = 131, P10 = 154, and P11 = 101.

  First, in step # 2 in FIG. 15A, the average value AVG is calculated based on the calculation formula of Formula 1 to be 127 (= (120 + 131 + 154 + 101) / 4). In the calculation of Equation 1, the decimal part is rounded off. In step # 3, the comparison result data Cij is C00 = C11 = 1 because P00 and P11 are smaller than the average value AVG (= 127), and P01 and P10 are larger than the average value AVG (= 127). C10 = 0 is calculated. Further, when the deviation value SD is calculated based on the calculation formula of Formula 2 in step # 4, 16 (= (| 120-127 | + | 131-127 | + | 154-127 | + | 101-127 |) / 4). In the calculation of Equation 2, the decimal part is rounded off. The encoded data includes an 8-bit average value AVG, 4-bit comparison result data Cij, and an 8-bit deviation value SD, and is a total of 20-bit data. Since one pixel block of 2 × 2 pixels before encoding is 32-bit data in total, the compression rate of the data amount is 62.5% (= 20/32 × 100).

  Next, in step # 7, 20-bit encoded data of each pixel block is read out, and in step # 8, from the read average value AVG (= 127), comparison result data Cij, and deviation value SD (= 16). When the 8-bit pixel value P′ij of each pixel block is decoded, P′00 = P′11 = 127−16 = 111 and P′01 = P′10 = 127 + 16 = 143 are obtained.

JP 2001-8206 A Japanese Patent No. 3696404 Japanese Patent No. 2616652 Japanese Patent No. 3770380

  As described above and as disclosed in Patent Document 1, the conventional BTC encoding processing method divides image data into a plurality of pixel blocks, and an average value AVG of each pixel value in the pixel block. The binarization process is performed by using the deviation value SD of each pixel value and the comparison result data Cij composed of “0” and “1”, and encoding is performed.

  However, in the conventional BTC encoding processing method, when the pixel values Pij of the pixel block are all the same as shown in FIG. 19A, the average value AVG becomes the same value as the pixel values Pij of the pixel block, The deviation value SD becomes zero. The comparison result data Cij are all 0, and each pixel value P′ij decoded based on the encoded data is the same as Pij in the original pixel block, that is, the average value AVG. In this case, in the process of decoding the encoded data, data other than the average value AVG is not effectively used, and the comparison result data Cij and the deviation value SD in the encoded data are unnecessarily stored in the encoded data. You are consuming memory capacity.

  As shown in FIG. 19B, when one of the pixel values Pij of the pixel block is different from the other pixel values by 1, for example, P00 = P10 = P11 = 114, P01 = 113, The average value AVG is equal to the pixel value 114, the deviation value SD is 0, and the comparison result data Cij is C00 = C10 = C11 = 0 and P01 = 1. The pixel values P′ij decoded based on the encoded data are all the same as the average value AVG as in the case where the pixel values Pij are all the same. Even when one of the pixel values Pij of the pixel block is different from the other pixel values by 1, the data other than the average value AVG is effectively used as in the case where each pixel value Pij is the same. In other words, the comparison result data Cij and the deviation value SD in the encoded data unnecessarily consume the memory capacity of the encoded data.

  In other words, in the conventional BTC encoding processing method, when all the pixel values Pij of the pixel block are the same or when the difference between the pixel values is not remarkable due to approximation to each other, the comparison result data in the encoded data Cij and deviation value SD unnecessarily consume the memory capacity of the encoded data. Therefore, if each pixel block obtained when the original image data is divided into pixel blocks includes a considerable amount of pixel blocks where the difference between the pixel values is not significant with respect to all the pixel blocks, the encoded data is stored. As a result, the memory capacity of the image memory is unnecessarily large and not used effectively.

  On the other hand, with regard to the image data processing device, an image memory for storing image data (gradation data) for one frame is prepared in the liquid crystal display device, and the gradation data of the frame to be displayed and the gradation data of the previous frame are stored. In comparison, an overdrive method has been proposed in which gradation data is corrected based on the comparison result, and processing is performed so as to reach a desired gradation data display at a desired response speed (for example, Patent Document 3 and Japanese Patent Application Laid-Open No. H10-228707). (See Patent Document 4).

  In the above-described overdrive method, image data for one frame is always required, and as the resolution increases, the capacity of the image memory that holds the image data inevitably increases. As a method for compressing the image data stored in the image memory and reducing the storage capacity of the image memory, the image memory is obtained by an existing compression method represented by the conventional BTC encoding processing method shown in Patent Document 2 above. A method has been devised to reduce the storage capacity of the disk. However, as described above, in the conventional BTC encoding processing method, the memory capacity of the image memory for storing the encoded data is unnecessarily used and not effectively used. There is a disadvantage that it cannot be used.

  The present invention has been made in view of the above problems, and an object of the present invention is to be generated when image data composed of pixel data having a predetermined number of pixels as a data holding unit is compressed by the BTC encoding method. Provided are an image encoding apparatus and method capable of further reducing the amount of encoded compressed data, and an image decoding apparatus and method for expanding the encoded compressed data. Second, the image encoding apparatus and image An object of the present invention is to provide an image data processing device using a decoding device.

  To achieve the above object, an image encoding apparatus according to the present invention divides image data composed of pixel data having a predetermined number of pixels as a data retention unit into unit blocks composed of a plurality of adjacent data retention units. Image data dividing means, an average value calculating means for calculating an average value of element values of each data holding unit in the unit block generated by the image data dividing means for each unit block, and the unit block A variation degree determination unit that determines the degree of variation in the distribution of element values of each data holding unit in the unit block based on a predetermined determination criterion and generates a flag bit indicating the determination result for each time And, for each unit block, each of the element value of each data holding unit in the unit block and the average value calculated by the average value calculation means Comparison result data generating means for generating comparison result data indicating a small comparison result, and the element value of each data holding unit in the unit block and the average value calculated by the average value calculating means for each unit block Deviation value calculating means for calculating an average deviation degree from the average value of the element value of each data holding unit in the unit block based on each difference value of the unit block; According to the value of the flag bit generated by the variation degree determination means, when the variation degree is small, the average value and the flag bit are combined to generate encoded data, and when the variation degree is large Encoded data generation means for generating encoded data by combining the average value, the flag bit, the comparison result data, and the deviation value. And butterflies.

  According to the image encoding device having the first feature, image data can be compressed by BTC encoding processing, and each element value (corresponding to a pixel block in a conventional BTC encoding processing method) (Corresponding to pixel values) are all the same or approximate to each other and the difference between the element values is not significant, the encoded data is a data structure that does not include the comparison result data and the deviation value. Since the data amount of the entire image data after the encoding process can be further compressed, it is possible to avoid unnecessarily consuming the memory capacity of the encoded data.

  Here, since the encoded data for each unit block by the image encoding device according to the present invention includes flag bits that are not included in the encoded data of the pixel block in the conventional BTC encoding processing method, the amount of data corresponding to that includes However, in normal image data, there are a considerable number of unit blocks in which the element values are all the same or close to each other and the difference between the element values is not significant, so the comparison result data and the deviation value are not included. Since the data reduction amount in the unit block greatly exceeds the data increase amount by the flag bit, the data amount can be compressed as the entire image data.

  In addition to the first feature, the image coding apparatus according to the present invention further includes: the variation degree determination unit includes: an element value of the data holding unit and the data value in all the data holding units in the unit block. When the absolute value of the difference value of the average value calculated by the average value calculation means is smaller than a predetermined determination threshold, it is determined that the variation degree is small, and otherwise, it is determined that the variation degree is large. This is the second feature.

  In addition to the first feature, the image coding apparatus according to the present invention further determines that the variation degree is small when the deviation value is smaller than a predetermined determination threshold, A third feature is that when the deviation value is equal to or greater than the predetermined determination threshold, it is determined that the degree of variation is large.

  According to the image encoding device of the second or third feature, the variation degree determination means determines whether or not the distribution degree of the element value distribution of each data holding unit in the unit block is large. Can be determined. That is, the difference value in the second feature indicates the variation degree of each element value, and the deviation value of the third feature indicates the average variation degree of each element value. The degree of variation can be determined.

  In addition to any one of the first to third features, the image encoding device according to the present invention further includes the encoded data generation means for all the unit blocks regardless of the value of the flag bit. Generating the first encoded data composed of the average value and the flag bit, and identifying whether the unit block has the large variation degree according to the value of the flag bit, A fourth feature is that second encoded data including the comparison result data and the deviation value is generated only for the large unit block.

  According to the image encoding device having the fourth feature, the entire encoded data can be separated into first encoded data and second encoded data and stored in an image memory or the like. Therefore, it is not necessary to handle as a set of encoded data having different code lengths for each unit block, and the configuration of an image memory or the like for storing the encoded data can be simplified. In addition, when the encoded data for each unit block is synthesized, the first encoded data is read, and only the second encoded data is additionally read and synthesized with the first encoded data only for the unit block having a large variation degree. As a result, encoded data can be obtained.

  In addition to the fourth feature, the image encoding device according to the present invention further includes an image memory for storing the encoded data, and a memory space of the image memory stores the first encoded data. A fifth feature is that the first area is divided into a second area for storing the second encoded data.

  According to the image coding apparatus having the fifth feature, the operational effects of the image coding apparatus having the fourth feature can be specifically realized.

  In addition to any of the above features, the image encoding device according to the present invention further includes the encoding data generation unit in a process in which the encoded data generation unit sequentially generates the encoded data for each unit block. When the accumulated data amount of data exceeds a predetermined upper limit value, the encoded data and the encoded data generated thereafter are generated without synthesizing the comparison result data and the deviation value, A sixth feature is that the determination criterion used by the variation degree determination unit is changed so that the variation degree is easily determined to be small.

  According to the image encoding device of the sixth feature, the data amount of the entire encoded data can be suppressed within a predetermined range, so that the encoded data of all the unit blocks is stored in an image memory having a fixed storage capacity, etc. In the case where encoded data is decoded in a later process, it is possible to prevent data from being lost. Furthermore, it is possible to more appropriately determine the degree of variation for all unit blocks under the condition that the data amount of the entire encoded data is limited for image data to be encoded next time. Therefore, it is possible to create encoded data in which comparison result data and a deviation value necessary for decoding reflecting the variation degree are added to a unit block having a relatively large variation degree.

  In addition to any of the above features, the image encoding device according to the present invention further includes 1 in one pixel when the pixel data is composed of gradation data of a plurality of different color components in one pixel. One color component is handled as the data holding unit, the image data dividing means, the average value calculating means, the variation degree determining means, the comparison result data generating means, the deviation value calculating means, and the encoded data generating means The seventh feature is that each processing is performed for each color component.

  According to the image coding apparatus of the seventh feature, the conventional BTC is compared with the image data in the case where the pixel data in the RGB format or the like is composed of each gradation data of a plurality of different color components in one pixel. Data compression that can reduce the amount of data after compression by the encoding processing method becomes possible.

  In the image encoding device according to the present invention, in addition to any one of the first to sixth features, the pixel data further includes luminance data in pixel units, pixel units or two adjacent pixels or four pixels. In the case of two types of color difference data in pixel units, for the luminance data, one pixel is handled as the data holding unit, and for the two types of color difference data, 1, 2 or 4 pixels are held in the data Treated as a unit, the image data dividing unit, the average value calculating unit, the variation degree determining unit, the comparison result data generating unit, the deviation value calculating unit, and the encoded data generating unit perform the respective processes. The eighth feature is that processing is performed separately for the luminance data and the two types of color difference data.

  According to the image coding apparatus of the eighth feature, pixel data in the YUV format, YUV422 format, YUV411 format, etc. is converted into luminance data in units of pixels and 2 in units of pixels or in units of two or four pixels adjacent to each other. Data compression that can reduce the amount of data after compression is possible for image data in the case of being composed of different types of color difference data, compared to the conventional BTC encoding processing method.

  In addition to the eighth feature, the image encoding device according to the present invention further includes original image data in which the pixel data is composed of gradation data for each color component in pixel units, and the pixel data is a pixel. According to a ninth feature of the present invention, there is provided image format conversion means for converting the luminance data in units and image data composed of two types of color difference data in units of pixels or in units of two or four pixels adjacent to each other.

  According to the image coding apparatus of the ninth feature, the original image data in the case where the pixel data in the RGB format or the like is composed of each gradation data of a plurality of different color components in one pixel is converted into image format conversion means. After the pixel data in the YUV422 format, YUV411 format, etc. is converted into image data when it is composed of luminance data in pixel units and two types of color difference data in units of two or four pixels adjacent to each other. Since the encoding process of the eighth feature is performed, the encoding process is performed after reducing the data amount of the color difference data, so that data compression that can further reduce the data amount after compression is possible.

  An image decoding apparatus according to the present invention decodes the encoded data encoded by the image encoding apparatus having any one of the first to sixth features into image data composed of pixel data of the data holding unit. An image decoding device that performs, for each unit block, the average value and the flag bit constituting the encoded data, or the average value, the flag bit, the comparison result data, and the deviation value, Data read means for individually reading, and a variation degree identification for identifying the degree of variation in the distribution of element values of each data holding unit in the unit block based on the value of the flag bit read by the data read means And the data reading means reads out the unit block identified as having a large degree of variation by the means and the degree of variation identifying means. A first decoding unit that generates decoding data based on the average value, the comparison result data, and the deviation value; and the unit block that is identified as having a small variation degree by the variation degree identifying unit. And a second decoding unit that generates decoded data based on the average value read by the data reading unit.

  According to the image decoding device of the first feature, the encoded data encoded by the image encoding device according to the present invention is an average value, comparison result data, and deviation value for a unit block having a large variation degree. Therefore, it is possible to generate the decoded data of each element value expanded to the same data length as the original pixel data by reflecting the variation of each element value by the first decoding means. For a unit block with a small degree, since the encoded data does not have the comparison result data and the deviation value, the variation of each element value by the second decoding means is small, so that the variation is ignored and based on the average value, Decoding data of each element value expanded to the same data length as the original pixel data can be generated and encoded by the image encoding device having any one of the first to sixth features Over the decoding of the data image decoding apparatus which is suitable for (data decompression) is provided.

  In addition to the first feature, the image decoding apparatus according to the present invention further includes: the first decoding unit calculates an element value of each data holding unit after decoding in the unit block; Based on the comparison result data corresponding to the data holding unit, it is calculated by adding or subtracting the deviation value with respect to the average value, and the second decoding unit is configured to perform decoding after the decoding in the unit block. A second feature is that all the element values of each data holding unit are set to the average value.

  According to the image decoding device of the second feature, the first decoding unit and the second decoding unit that achieve the operational effects of the image decoding device of the first feature are realized concretely and simply. be able to.

  In addition to the first or second feature, the image decoding device according to the present invention further includes a first data reading unit that reads the average value and the flag bit from the encoded data. A third feature is that the apparatus further comprises second data reading means for reading the comparison result data and the deviation value from the encoded data when the degree of variation is identified by the variation degree identifying means. And

  According to the image decoding apparatus of the third feature, since the data reading unit is divided into the first data reading unit and the second data reading unit, the first data reading unit is a unit block to be read. The average value and flag bits of the encoded data can be read regardless of the degree of variation of the data, and the second data reading means can detect the deviation from the comparison result data of the encoded data only for the unit block having a large degree of variation. Since the value can be read out, the image encoding device stores the average value and flag bits of all unit blocks in a single storage area, and the comparison result data and deviation value of the unit block having a large degree of variation are stored separately. When storing in one storage area, the average value and flag bit, and the comparison result data and deviation value are separately separated from each storage area. Data reading means for reading a manner can be realized. Further, since the data width of the encoded data read by the first data reading unit and the data width of the encoded data read by the second data reading unit are fixed, the hardware of the first data reading unit and the second data reading unit is fixed. The hardware configuration can be simplified.

  An image data processing device according to the present invention includes an image encoding device having any one of the first to fourth features, an image memory for storing encoded data encoded by the image encoding device, and the image memory And the image decoding device having any one of the first to third features for decoding the encoded data stored in the device.

  According to the image data processing apparatus having the above characteristics, it is possible to handle data stored by compressing the data amount as compared with the conventional BTC encoding processing method as image data related to image display, and temporarily storing the image data It is possible to reduce the memory capacity required for the operation.

  An image encoding method according to the present invention includes an image data dividing step of dividing image data composed of pixel data having a predetermined number of pixels as a data retention unit into unit blocks composed of a plurality of adjacent data retention units; For each unit block, an average value calculating step for calculating an average value of element values of each data holding unit in the unit block generated in the image data dividing step; and for each unit block, in the unit block A variation degree determination step for determining a degree of variation in the distribution of element values of each data holding unit based on a predetermined determination criterion and generating a flag bit indicating the determination result; and for each unit block A comparison result data indicating the result of comparing the magnitude of the element value of each data holding unit in the unit block and the average value calculated in the average value calculating step. A comparison result data generation step for generating data, and for each unit block, the element value of each data holding unit in the unit block and the difference value between the average values calculated in the average value calculation step A deviation value calculating step for calculating a deviation value indicating an average deviation degree from the average value of the element values of the respective data holding units in the unit block, and the variation degree determining step for each unit block. Depending on the value of the flag bit, when the degree of variation is small, the average value and the flag bit are combined to generate encoded data, and when the degree of variation is large, the average value and the flag And an encoded data generation step of generating encoded data by combining the bit, the comparison result data, and the deviation value.

  According to the image encoding method of the above feature, when image data can be compressed by BTC encoding processing, and when the element values of the unit blocks are all the same or close to each other and the difference between the element values is not remarkable Since the encoded data has a data structure that does not include the comparison result data and the deviation value, in this case, the data amount of the entire image data after the encoding process can be further compressed. It is possible to avoid consuming data memory capacity.

  Here, since the encoded data for each unit block according to the image encoding method according to the present invention includes flag bits that are not included in the encoded data of the pixel block in the conventional BTC encoding processing method, the amount of data corresponding to that includes However, in normal image data, there are a considerable number of unit blocks in which the element values are all the same or close to each other and the difference between the element values is not significant, so the comparison result data and the deviation value are not included. Since the data reduction amount in the unit block greatly exceeds the data increase amount by the flag bit, the data amount can be compressed as the entire image data.

  An image decoding method according to the present invention is an image decoding method for decoding the encoded data encoded by the image encoding method having the above characteristics into a data format before encoding, for each unit block. The average value and the flag bits constituting the encoded data are individually read out from the first data reading step, and the flag bits read out in the data reading step are used to determine the respective values in the unit block. A variation degree identifying step for identifying the degree of variation in the distribution of element values of the data holding unit, and the comparison result constituting the encoded data when the variation degree is identified as being large in the variation degree identifying step In the data second reading step for reading out the data and the deviation value individually and in the variation degree identifying step, it is identified that the variation degree is large. A first decoding step for generating decoded data based on the average value read in the first data reading step, the comparison result data read in the second data reading step, and the deviation value; And a second decoding step for generating decoded data based on the average value read in the first data reading step when the variation degree is identified as being small in the variation degree identifying step. It is characterized by that.

  According to the image decoding method of the above feature, the encoded data encoded by the image encoding method according to the present invention includes an average value, comparison result data, and a deviation value for a unit block having a large variation degree. The average value, the comparison result data, and the deviation value are read in the first and second data reading steps, and the first decoding step is expanded to the same data length as the original pixel data, reflecting the variation of each element value. Decoded data of each element value can be generated, and for a unit block with a small variation degree, the encoded data does not include the comparison result data and the deviation value. Therefore, in the second decoding step, each element value Since the variation is small, each element value expanded to the same data length as the original pixel data based on the average value read in the first data reading step ignoring the variation. It can generate-coding data, an image decoding method which are suitable for coded decoding of coded data in the image coding method according to the present invention (data decompression) is provided.

  Next, the image coding apparatus and method according to the present invention (hereinafter referred to as “the present invention coding apparatus” and “the present invention coding method” as appropriate), and the image decoding apparatus and method according to the present invention (hereinafter referred to as appropriate). The image data processing apparatus including the “encoding apparatus of the present invention” and the “decoding method of the present invention”, and the encoding apparatus and the decoding apparatus of the present invention will be described with reference to the drawings.

<First Embodiment>
First, a schematic system configuration of the encoding apparatus and decoding apparatus of the present invention will be described. As shown in FIG. 1, the encoding apparatus 10 of the present invention includes an image data dividing unit 11, an average value calculating unit 12, a variation degree determining unit 13, a comparison result data generating unit 14, a deviation value calculating unit 15, and an encoded data generating unit. The decoding device 20 of the present invention includes a data reading unit 21, a variation degree identifying unit 22, a first decoding unit 23, and a second decoding unit 24. Configured.

  The image data dividing unit 11 divides image data composed of pixel data having a predetermined number of pixels (for example, 1, 2 or 4 pixels) as a data retention unit into unit blocks composed of a plurality of adjacent data retention units. . In the first embodiment, as the image data to be encoded, as shown in FIG. 2, the resolution is VGA (X direction: 640 pixels, Y direction: 480 pixels), and the color display format is RGB format (color 3). Primary color, R: red, G: green, B: blue, each color data is displayed), and the gradation is 8-bit RGB data of 16.77 million colors, and each pixel's RGB components are data holding units. Assume that the image data is composed of pixel data. Hereinafter, for convenience of description, image data to be encoded may be referred to as “primary image data” as appropriate in order to distinguish it from image data after encoding.

  In addition, the unit block corresponds to an N × N pixel block (usually N = 2 or 4) of the conventional BTC encoding method. In the first embodiment, as shown in FIG. A pixel block composed of 2 × 2 pixels is used. Therefore, the RGB components of the primary image data are divided by the image data dividing unit 11 into unit blocks of 320 blocks in the X direction and 240 blocks in the Y direction. It should be noted that the element values (8-bit data of RGB gradation values 0 to 255) constituting the unit block of each RGB component are appropriately set to Pij (i = 0, 1, j) as shown in FIG. = 0, 1).

  The average value calculation means 12 calculates the average value AVG of the four element values Pij of each unit block based on the following equation 3 as in the conventional BTC encoding method, For temporary storage. In the first embodiment, since each element value Pij is 8-bit data, the average value AVG is also 8-bit data.

  The variation degree determination means 13 determines the degree of variation of the distribution of the four element values Pij of each unit block based on a predetermined criterion, and generates a 1-bit flag bit FLG indicating the determination result. . In the first embodiment, for each unit block, a magnitude comparison determination is made as to whether the absolute value (| Pij−AVG |) of the difference between each element value Pij and the average value AVG is greater than or less than a predetermined threshold Dth. When the absolute value (| Pij−AVG |) is equal to or greater than the threshold value Dth in at least one element value of the four pixels in the unit block, it is determined that the degree of variation is large, and the flag bit FLG Is set to “1”. If the absolute value (| Pij−AVG |) is less than the threshold value Dth among all the element values of the four pixels in the unit block, it is determined that the degree of variation is small, and the flag bit FLG is set to “0”. Set to "".

  The comparison result data generating means 14 generates 4-bit comparison result data indicating the respective magnitude comparison results of the four element values Pij and the average value AVG in the unit block for each unit block. The size comparison result of each unit block corresponds to each element value Pij on a one-to-one basis and is represented by Cij (i = 0, 1, j = 0, 1). Here, Cij = 0 when Pij> AVG, and Cij = 1 when Pij ≦ AVG.

  The deviation value calculation means 15 is, for each unit block, a deviation value indicating an average degree of deviation from the average value AVG of each element value Pij based on each difference value between the four element values Pij in the unit block and the average value AVG. The SD is calculated based on the following equation 4 and stored temporarily for later processing, as in the conventional BTC encoding method. In the first embodiment, since each element value Pij is 8-bit data, the deviation value SD is also 8-bit data.

  The encoded data generation means 16 synthesizes the average value AVG and the flag bit FLG and encodes 9 bits when the degree of variation is small (FLG = 0) according to the value of the flag bit FLG for each unit block. When the data ECD9 (corresponding to the first encoded data) is generated and the degree of variation is large (FLG = 1), the average value AVG, the flag bit FLG, the comparison result data Cij, and the deviation value SD are combined to generate 21 bits. Encoded data ECD21 is generated and stored in the image memory 17, respectively. If the comparison result data Cij and the deviation value SD are regarded as 12-bit partial encoded data ECD12 (corresponding to the second encoded data), the encoded data ECD21 has 9-bit encoded data ECD9 and a 12-bit partial. The encoded data ECD12 is synthesized. FIG. 4 schematically shows the data structure of the encoded data ECD9 and the encoded data ECD21.

  The image memory 17 is a storage device or storage means for storing the encoded data ECD9 and ECD21, and is configured as a dedicated memory for storing the encoded data, or used for storing other data. The general-purpose memory has a partial storage area for storing encoded data. In the first embodiment, the image memory 17 is a configuration attached to the encoding device 10 of the present invention. However, a configuration attached to the decoding device 20 of the present invention, or the encoding device 10 of the present invention and the present invention. The configuration may be provided independently of the decoding device 20.

  Depending on the configuration of the image memory 17, there are two possible methods for storing the encoded data ECD 9 or ECD 21 in the image memory 17. Note that the procedure for sequentially selecting a total of 76800 (= 320 × 240) unit blocks of 320 in the X direction and 240 in the Y direction starts, for example, in order from the unit block located in the upper left corner, and + X direction (right direction) Scan 320 blocks, move one step at a time in the -Y direction (downward), repeat the scan in the right direction 240 times, finally select the unit block in the lower right corner, and finish.

  As shown in FIG. 5, the first storage method of the encoded data in the image memory 17 includes a first area having a 9-bit data width for storing the 9-bit encoded data ECD9. This is a storage method specialized in the case of being divided into a second area having a 12-bit data width for storing the 12-bit partially encoded data ECD12. In this case, the encoded data generation means 16 generates encoded data ECD9 for each unit block that is sequentially selected, regardless of the value of the flag bit FLG, and stores the encoded data ECD9 in the first area of 9-bit width in the image memory 17. Store in parallel. After being stored in the first area of the image memory 17, the partial encoded data ECD 12 is generated only when the value of the flag bit FLG is “1”, and the partial encoded data ECD 12 is generated in parallel with the second area of the image memory 17. In addition, the partial encoded data ECD12 of the unit block whose value of the previous flag bit FLG is “1” is stored successively. Here, the address depth of the first area of the image memory 17 is prepared for all unit blocks (76800), whereas the address depth of the second area is, for example, the address of the first area. Half the depth (38400) is provided. That is, based on an empirical rule that unit blocks (FLG = 1) having a large variation degree of each element value Pij among all unit blocks can be suppressed to half or less by appropriately setting the threshold value Dth. Yes. Accordingly, the address depth of the second area changes according to the set value of the threshold value Dth.

  The second storage method of the encoded data in the image memory 17 is a storage method specialized when the image memory 17 is composed of a 1-bit wide FIFO (First In First Out) memory. In this case, the encoded data generation means 16 generates encoded data ECD9 for each unit block selected sequentially, regardless of the value of the flag bit FLG, and stores the average value AVG and the flag bit FLG in the image memory 17. Only when the value of the flag bit FLG is “1”, the partial encoded data ECD12 is continuously generated and stored bit by bit after the encoded data ECD9. Alternatively, when the value of the flag bit FLG is “0”, the encoded data generation unit 16 generates the encoded data ECD9 and stores the encoded data ECD9 in the image memory 17 bit by bit in the order of the average value AVG and the flag bit FLG. When the value of the flag bit FLG is “1”, the encoded data ECD21 is generated, and the portion of the first half of the encoded data ECD9 is stored one bit at a time in the order of the average value AVG and the flag bit FLG. Partially encoded data ECD12 is stored bit by bit. Here, the storage capacity of the image memory 17 is 76800 × 9 bits for storing 9-bit encoded data ECD9 of all unit blocks (76800), and 12 bits of half of the entire unit block (38400). The total number of bits (1152000) of 38400 × 12 bits for storing the partially encoded data ECD12 is prepared for each RGB component. That is, among all the unit blocks, the unit block (FLG = 1) having a large variation degree of each element value Pij can be suppressed to less than half by appropriately setting the threshold value Dth, as will be described later. Based on rules of thumb. Accordingly, the storage capacity of the image memory 17 changes according to the set value of the threshold value Dth.

  The data reading means 21 is configured to read the encoded data ECD9 or ECD21 stored in the image memory 17 for each unit block. Specifically, the data reading unit 21 includes a first data reading unit 21a that reads the encoded data ECD9 and a second data reading unit 21b that reads the partially encoded data ECD12.

  The variation degree identifying unit 22 identifies the degree of variation in the distribution of the element values Pij in the unit block based on the value of the flag bit FLG of the encoded data ECD9 read by the first data reading unit 21a.

  The method by which the data reading means 21 reads the encoded data ECD9 or ECD21 from the image memory 17 depends on the configuration of the image memory 17 as well as the method for storing the encoded data ECD9 or ECD21 in the image memory 17. A street method is conceivable.

  The first method for reading the encoded data from the image memory 17 corresponds to the first method for storing the encoded data in the image memory 17 described above, and the image memory 17 stores the encoded data of 9 bits. Specially, it is divided into a first area having a 9-bit data width for storing ECD9 and a second area having a 12-bit data width for storing 12-bit partially encoded data ECD12. This is a generalized reading method. In this case, the first data reading means 21a converts the 9-bit encoded data ECD9 from the 9-bit wide first area of the image memory 17 to each unit block sequentially selected regardless of the value of the flag bit FLG. Read in bit parallel. Then, only when the value of the flag bit FLG in the read encoded data ECD9 is “1”, the second data reading means 21b performs the 12-bit partial code from the 12-bit wide second area of the image memory 17. The data ECD12 is read out in parallel with each bit.

  Here, since the encoded data ECD9 of all the unit blocks is stored in the first area of the image memory 17 in the scanning order of the unit blocks, reading the corresponding unit block in the same order in the scanning order of the unit blocks. The encoded data ECD9 is read correctly. In the second area of the image memory 17, only the partial encoded data ECD12 of the unit block having a large variation degree with the value of the flag bit FLG being “1” is continuously stored in the unit block scanning order. That is, since the storage area of the unit block having a small variation degree with the flag bit FLG value of “0” is stored without being vacated, the partial encoded data ECD12 of the corresponding unit block is sequentially assigned to each unit block. Will be read correctly.

  The second method for reading the encoded data from the image memory 17 corresponds to the second method for storing the encoded data in the image memory 17 described above, and the image memory 17 is a FIFO memory having a 1-bit width. This is a reading method specialized for the case where In this case, the first data reading means 21a serially reads the 9-bit encoded data ECD9 from the image memory 17 bit by bit for each unit block selected sequentially, regardless of the value of the flag bit FLG. Then, only when the value of the flag bit FLG in the read encoded data ECD9 is “1”, the second data reading means 21b continues to output the 12-bit partial encoded data ECD12 from the image memory 17 bit by bit. Read serially.

  Here, since the encoded data ECD9 or ECD21 of all the unit blocks is stored in the image memory 17 in the scanning order of the unit blocks, the encoded data ECD9 or ECD21 of all the unit blocks is similarly stored in the scanning order of the unit blocks. By reading the encoded data ECD9 located in the head part in common, the encoded data ECD9 of the corresponding unit block is read correctly. Further, since the encoded data ECD21 of the unit block having a large variation degree with the flag bit FLG value “1” has the partial encoded data ECD12 added after the encoded data ECD9, the value of the flag bit FLG is “ In the case of 1 ″, the encoded data ECD21 of the unit block can be read. When the value of the flag bit FLG is “0”, the encoded data ECD9 or ECD21 of the unit block in the next operation order can be read. In other words, by checking the value of the flag bit FLG, whether the encoded data of the unit block is completed with the encoded data ECD9 read out first, or the partial encoded data ECD12 remains unread after that. You can see if there is.

  The first decoding unit 23 differs from the average value AVG of the encoded data ECD21 read by the data reading unit 21 and the comparison result data Cij with respect to the unit block identified as having a large variation by the variation degree identifying unit 22. Based on the value SD, decrypted data P′ij (i = 0, 1, j = 0, 1) is generated. The decoded data P′ij represents the element value after decoding of the unit block, and corresponds to the element value Pij before encoding. Specifically, as in the conventional BTC encoding method, when comparison result data Cij is “0”, P′ij = AVG + SD, and when comparison result data Cij is “1”, P′ij = AVG−SD. And

  The second decoding unit 24 decodes the decoded data P based on the average value AVG of the encoded data ECD9 read by the data reading unit 21 for the unit block identified by the variation degree identifying unit 22 as having a small variation degree. 'ij (i = 0, 1, j = 0, 1) is generated. Specifically, unlike the conventional BTC encoding method, since the encoded data ECD9 does not include the comparison result data Cij and the deviation value SD, these decoded data P ′ are not used without using these data. Let ij be the average value AVG uniformly.

  The decoded data P′ij decoded by the first decoding unit 23 or the second decoding unit 24 has the same data format and data length (8-bit RGB data) as the element value Pij of the unit block before encoding. It is.

  As described above, each of the means 11 to 16 constituting the encoding apparatus 10 of the present invention and each of the means 21 to 24 constituting the decoding apparatus 20 of the present invention have been described in detail. Configuration by software means for realizing each function by execution of a computer program on a computer, configuration by hardware means for realizing each function by hardware, and configuration by combining software means and hardware means Any configuration may be used.

  Next, the procedure in which the encoding apparatus 10 of the present invention executes the BTC encoding process (data compression process) according to the encoding method of the present invention will be described with reference to FIG. FIG. 6 is a flowchart showing a schematic processing procedure of the encoding method of the present invention.

  First, when the encoding apparatus 10 of the present invention receives primary image data (for example, VGA, RGB format data), the image data dividing means 11 is a unit composed of a plurality of data holding units adjacent to each other. Divide into blocks (image data dividing step: step # 11). As a unit block, for example, a pixel block composed of 2 × 2 pixels shown in FIG. 3 is used.

  Next, the average value calculation means 12 calculates the average value AVG of the four element values Pij of one unit block selected in order by the above equation 3 (average value calculation step: step # 12). In the calculation of Equation 3, the decimal part is rounded off.

  Subsequently, the variation degree determination means 13 determines the degree of variation of the distribution of the four element values Pij in the selected unit block in the above-described manner, and sets the flag bit FLG based on the determination result. “0” or “1” is set (variation degree determination step: step # 13).

  When “1” is set in the flag bit FLG in the above step # 13, the comparison result data generating means 14 displays the respective size comparison results of the four element values Pij and the average value AVG in the selected unit block. The 4-bit comparison result data Cij shown is generated as described above (comparison result data generation step: step # 14).

  Subsequent to step # 14, the deviation value calculation means 15 causes the deviation indicating the average degree of deviation from the average value AVG of each element value Pij based on the difference values between the four element values Pij and the average value AVG in the unit block. The value SD is calculated based on the equation (4) (deviation value calculating step: step # 15). In the calculation of Equation 4, the decimal part is rounded off.

  Subsequent to step # 15, the encoded data generating means 16 combines the average value AVG, the flag bit FLG, the comparison result data Cij, and the deviation value SD calculated in steps # 12 to # 15 to generate 21-bit encoded data ECD21. Is generated and stored in the image memory 17 (encoded data generation step 1: step # 16).

  When “0” is set to the flag bit FLG in step # 13, the comparison result data Cij is not calculated by the comparison result data generating unit 14 and the deviation value SD is not calculated by the deviation value calculating unit 15. The encoded data generating means 16 generates the 9-bit encoded data ECD9 by combining the average value AVG calculated in steps # 12 and # 13 and the flag bit FLG, and stores it in the image memory 17 (encoded data). Generation process 2: Step # 17).

  In the encoded data generation steps 1 and 2, the encoded data ECD21 and ECD9 are stored in the image memory 17 by any one of the above-described first and second storage methods, and therefore redundant description is omitted.

  When the encoded data generation process of step # 16 or # 17 is completed for one unit block, it is determined whether the completed unit block is the last unit block (step # 18). Selects the next unit block (step # 19), and repeats the processing from steps # 12 to # 18 until it is determined as the last unit block. In addition, each process of step # 12- # 17 is performed 3 times about each RGB component of the same unit block.

  Next, the procedure for the decoding apparatus 20 of the present invention to execute the BTC decoding process (data expansion process) according to the decoding method of the present invention will be described with reference to FIG. FIG. 7 is a flowchart showing a schematic processing procedure of the decoding method of the present invention.

  First, the first data reading means 21a obtains the 8-bit average value AVG and the 1-bit flag bit FLG constituting the encoded data ECD9 of the encoded data ECD9 or ECD21 of one unit block selected in order. Then, it reads out from the image memory 17 (first data reading step: step # 21). The method of reading the encoded data ECD9 is performed by the first or second reading method according to the configuration of the image memory 17.

  Subsequently, when the value of the flag bit FLG is “1” based on the value of the flag bit FLG read by the step # 21, the variation degree identifying unit 22 varies the distribution of the element values Pij in the unit block. On the contrary, when the value of the flag bit FLG is “0”, it is identified that the variation degree is small (variation degree identification step: step # 22).

  If it is determined in step # 22 that the degree of variation is large (FLG = 1), the second data reading unit 21b performs 4-bit comparison result data Cij (i = i) constituting the partial encoded data ECD12 of the encoded data ECD21. 0, 1, j = 0, 1) and the 8-bit deviation value SD are individually read out (second data reading step: step # 23). The method of reading the partially encoded data ECD 12 is performed by the first or second reading method according to the configuration of the image memory 17.

  Subsequent to step # 23, the first decoding unit 23 determines each value in the unit block based on the average value AVG read in step # 21, the comparison result data Cij read in step # 23, and the deviation value SD. Decoded element data P′ij is generated and output to the image display device that displays and outputs the decoded data P′ij (first decoding step: step # 24). Specifically, as in the conventional BTC encoding method, when comparison result data Cij is “0”, P′ij = AVG + SD, and when comparison result data Cij is “1”, P′ij = AVG−SD. And

  If it is determined in step # 22 that the degree of variation is small (FLG = 0), the second data reading means 21b does not read the comparison result data Cij and the deviation value SD, and the second decoding means 24 The decoded data P′ij of the four element values in the unit block is uniformly generated as the average value AVG read in step # 21, and the decoded data P′ij is output to the image display device (second). Decryption step: Step # 25).

  When the first or second decoding step (step # 24 or # 25) is completed for one unit block, it is determined whether the completed unit block is the last unit block (step # 26). If it is not a unit block, the next unit block is selected (step # 27), and the processing from steps # 21 to # 26 is repeated until it is determined as the last unit block. In addition, each process of step # 21- # 25 is performed 3 times about RGB component of the same unit block.

  FIG. 8 shows a specific example of processing results of the encoding method and decoding method of the present invention in one unit block.

  When the element value Pij for 256 gradations of the unit block is P00 = 120, P10 = 131, P01 = 154, P11 = 101, and the threshold Dth is set to 13, (A) and 29 are set. The case (B) will be described. The average value AVG is calculated as 127 regardless of the threshold value Dth. If the absolute value | Pij−AVG | of the difference between the element value Pij and the average value AVG is calculated in order to determine the degree of variation, (7, 5, 28, 26) is obtained. When the threshold value Dth is set to 13, (A) is equal to or greater than the threshold value Dth with two element values, it is determined that the degree of variation is large, and the flag bit FLG is set to “1”. On the other hand, when the threshold value Dth is set to 29 (B), it is determined that all the element values are less than the threshold value Dth, the degree of variation is small, and the flag bit FLG is set to “0”. In any case, the comparison result data Cij is (1, 0, 0, 1), and the deviation value SD is 16. However, when the threshold value Dth is set to 13, since the flag bit FLG is “1”, 21-bit encoded data ECD21 is generated and stored in the image memory. In the decoding process, when the decoded data P′ij is “0”, the comparison result data Cij is P′ij = AVG + SD, and when the comparison result data Cij is “1”, P′ij = AVG−SD. Therefore, P′00 = 111, P′10 = 143, P′01 = 143, and P′11 = 111 are decoded. On the other hand, when the threshold value Dth is set to 29 (B), since the flag bit FLG is “0”, 9-bit encoded data ECD9 is generated. In the decoding process, the decoded data P′ij is uniformly decoded into the average value AVG (= 127).

  Next, the data compression rate from the primary image data of the image data encoded by the encoding device 10 of the present invention will be examined.

  As in the above description, assuming that the above-described VGA data of 16.77 million colors of 8 bits for each of RGB is assumed as the primary image data, the data amount of the primary image data is (640 × 480) pixels × 24 bits = 7372800 bits. It is. When primary image data having the same data amount is encoded by the conventional BTC encoding method, the unit block (2 × 2 pixels, 32 bits) per RGB component is an 8-bit average. Since the value AVG and the 4-bit comparison result data Cij and the 8-bit deviation value SD are converted into encoded data of 20 bits in total, the data compression rate is 62.5% (= 20/32 × 100). The amount of data is 4608000 bits.

  On the other hand, the image data encoded by the encoding apparatus 10 of the present invention is 9 bits or 21 bits depending on the degree of variation of each element value Pij of the unit block (2 × 2 pixels, 32 bits). The encoded data is converted into the encoded data ECD9 or ECD21. Therefore, the code length is not constant. When the degree of variation of all the unit blocks is large, the data compression rate is 65.625% (= 21/32 × 100) and the total data amount is 4838400 bits, which is compressed by the encoding process using the conventional BTC encoding method. On the contrary, when the degree of variation of all unit blocks is small, the data compression rate is 28.125% (= 9/32 × 100), and the total data amount is 2073600 bits. Compared with the conventional BTC encoding method, the compression rate is significantly reduced, and the amount of data after compression can be reduced. Therefore, if the degree of variation is small in unit blocks exceeding about 8.34% (19200 blocks) of all unit blocks, the compression rate is lower than the data compression rate (62.5%) by the conventional BTC encoding method.

  When the primary image data is data of a natural image such as a photographic image or a moving image, if the unit block size is small, the difference between each element value Pij and its average value AVG is obtained for all elements in the unit block. It is considered that the absolute value (| Pij−AVG |) is likely to be suppressed to a small value. That is, when the unit block size is 2 × 2 pixels, the four element values Pij in the unit block are almost the same value, and the existence ratio of the unit block whose absolute value of the difference is less than the predetermined threshold Dth is increased. It is expected.

  As a result of verification using several photographic images, when the threshold value Dth is set to 13 (element value Pij is 0 to 255), about 70% of all unit blocks have an absolute value of the difference. It was determined that the threshold value was less than the threshold value Dth (= 13), that is, the degree of variation was small, and the data compression rate of the entire image data was approximately 40%. In the conventional BTC encoding method, the data compression rate was 62.5%, so it can be seen that the compression rate can be reduced by 20% or more.

  Looking at the verification results in more detail, in some of the above photographic image examples, the generation ratio of 9-bit encoded data ECD9 and 21-bit encoded data ECD21 is approximately 7: 3. Assuming that the generation ratio of the encoded data ECD21 is 50% or less with respect to all the unit blocks, and if the storage capacity of the image memory is 1152000 bits for each RGB component, all the unit blocks The two encoded data ECD9 and ECD21 can be reliably stored.

Second Embodiment
Next, a second embodiment of the encoding apparatus and method of the present invention and the decoding apparatus and method of the present invention will be described with reference to FIGS.

  In the first embodiment, the data of the RGB format is assumed as the primary image data, but the YUV format (Y: luminance, U, V: color difference) is used as the color display format in addition to the RGB format. Similarly, it is possible to perform the encoding process by the encoding method of the present invention and the decoding process by the decoding method of the present invention on the element value Pij of each component. is there.

  Furthermore, human visual characteristics are sensitive to luminance, but are insensitive to color differences, so that each pixel has luminance data (Y) and two types of color difference data (U, V). For the YUV format (YUV 444) (based on RGB data), two types of color difference data (U, V) are converted into two pixels adjacent in the horizontal direction or two pixels (2 × Image data can be compressed by providing one for every two pixels. In this case, one element value is created by thinning out or averaging one or three pixels out of two or four pixels.

  As such a compression type YUV format, there are a YUV422 format, a YUV411 format, and the like. FIG. 9 is a diagram schematically showing the data configuration of the YUV422 format and YUV411 format corresponding to RGB data for 4 × 4 pixels, and the data amount of each component in each YUV format and the data compression rate from the RGB format. Is shown.

  Equations 5 and 6 below show conversion equations between components between the RGB format and the uncompressed YUV format (YUV444).

  In the second embodiment, after receiving RGB data as shown in FIG. 2 as input image data, after converting to YUV format to generate primary image data, the primary image data in the YUV format is generated. Then, the encoding process according to the encoding method of the present invention is performed in the same processing procedure as in the first embodiment, and the encoded data ECD9 or ECD21 is generated for each unit block and stored in the image memory.

  In the second embodiment, the encoded data ECD9 or ECD21 for each unit block stored in the image memory is subjected to a decoding process by the decoding method of the present invention in the same processing procedure as in the first embodiment. Then, YUV-format decoded data P′ij corresponding to each element value Pij of the unit block before encoding is generated, and the YUV-format decoded data P′ij is inversely converted into the RGB format, and the image display device Output to.

  Therefore, the encoding apparatus 30 according to the second embodiment includes an image data dividing unit 11, an average value calculating unit 12, a variation degree determining unit 13, a comparison result data generating unit 14, a deviation value, as shown in FIG. In addition to the calculation unit 15, the encoded data generation unit 16, and the image memory 17, a format conversion unit 31 that converts the color display format as the preprocessing unit from the RGB format to the YUV format is provided. Furthermore, the decoding apparatus 40 according to the second embodiment includes a data reading unit 21, a variation degree identifying unit 22, a first decoding unit 23, and a second decoding unit 24 as shown in FIG. In addition, it comprises a format reverse conversion means 41 that reversely converts the color display format as post-processing means from the YUV format to the RGB format. Since the encoding device 30 of the present invention is the same as that of the first embodiment except for the format conversion means 31, the overlapping description is omitted. Since the decoding apparatus 40 of the present invention is the same as that of the first embodiment except for the format reverse conversion means 41, the duplicate description is omitted in the same manner.

  In the following description, the second embodiment will describe a case where a compression type YUV411 format is used as the YUV format.

  The format conversion means 31 calculates the average value of the color difference data U and V for 2 × 2 pixels after calculating the YUV component values of the RGB component values of each pixel based on the conversion formula of the above formula 5. Each element value. Instead of calculating the average value of the color difference data U and V, the color difference data U and V of any one pixel of 2 × 2 pixels may be extracted as a representative value. As described above, the primary image data in the YUV format generated by the format conversion means 31 is 640 × 480 × 8 bits for luminance data (Y), and 320 × 240 × 8 for two types of color difference data (U, V). The amount of data is in bits. For the color difference data (U, V), 2 × 2 pixels are one data holding unit.

  The format reverse conversion means 41 uses the element values P′ij (2 × 2 pixels) of the decoded color difference data (U, V) as the color difference data (U, V) of each pixel to decode the luminance Together with each element value P′ij (for one pixel) of data (Y), an RGB component value is calculated based on the conversion formula of Equation 6 above. In the first embodiment, the first decoding unit 23 and the second decoding unit 24 output the decoded data P′ij as it is to the image display device. However, in the second embodiment, Since the format reverse conversion process to the RGB format is involved, the first decoding means 23 and the second decoding means 24 do not output the decoded data P′ij to the image display device, but to the format reverse conversion means 41. The format reverse conversion means 41 outputs the RGB format decoded data after the format reverse conversion processing to the image display device.

  The format conversion means 31 and the format reverse conversion means 41 added in the second embodiment are configured by software means for realizing each function by executing a computer program on a microprocessor or a microcomputer, or each function. It is possible to adopt any configuration such as a configuration using hardware means for realizing the above in hardware, or a combination of software means and hardware means.

  Next, the procedure in which the encoding apparatus 30 of the present invention executes the BTC encoding process (data compression process) according to the encoding method of the present invention will be described with reference to FIG. FIG. 11 is a flowchart showing a schematic processing procedure of the encoding method of the present invention according to the second embodiment.

  First, when the encoding apparatus 30 of the present invention receives input image data (VGA, RGB format data), the format conversion means 31 converts it into primary image data in the YUV411 format (format conversion step: step # 10).

  Subsequently, the image data dividing unit 11 divides the primary image data into unit blocks composed of a plurality of data holding units adjacent to each other (image data dividing step: step # 11). The unit block uses a pixel block composed of 2 × 2 pixels for luminance data (Y), and is composed of a 2 × 2 data holding unit (4 × 4 pixels) for color difference data (U, V). Pixel block to be used. Therefore, the total number of unit blocks is 320 × 240 blocks for the luminance data (Y) and 160 × 120 blocks for the color difference data (U, V).

  Hereinafter, for each component of the luminance data (Y) and the color difference data (U, V), the processing from steps # 12 to # 18 is performed for the luminance data (Y, as in the processing procedure of the first embodiment shown in FIG. ) And all unit blocks of the two color difference data (U, V). As the threshold value Dth used in the variation degree determination step (step # 13), for example, “13” similar to that in the first embodiment is used. As the threshold value Dth, it is preferable to separately use values individually verified by simulation or the like for each of the luminance data (Y) and the color difference data (U, V).

  Next, the procedure for the decoding apparatus 40 of the present invention to execute the BTC decoding process (data expansion process) by the decoding method of the present invention will be described with reference to FIG. FIG. 12 is a flowchart showing a schematic processing procedure of the decoding method of the present invention.

  Similar to the processing procedure of the first embodiment shown in FIG. 7, the processing from step # 21 to # 25 is executed for the selected unit block of each component of the YUV411 data. However, in the first or second decoding step (step # 24 or # 25), the decoded data P'ij is not output to the image display device.

  When the luminance data (Y) and the two color difference data (U, V) decoded data P′ij corresponding to the same 2 × 2 pixels are generated, the format inverse conversion means 41 performs the decoded color difference data (U , V) each element value P′ij (2 × 2 pixels) as color difference data (U, V) of each pixel, and each element value P′ij (for one pixel) of decoded luminance data (Y) ) And RGB component values are calculated based on the above equation 6 and the RGB format decoded data after format reverse conversion processing is output to the image display device (format reverse conversion step: step # 28). Each unit block of the two color difference data (U, V) corresponds to four unit blocks of the luminance data (Y), and each of the four element values obtained by decoding the two color difference data (U, V). P′ij (one element value corresponds to 2 × 2 pixels) corresponds to one of four unit blocks of luminance data (Y). Referring to the YUV411 format diagram of FIG. 9, four element values (Y00, Y01, Y10, Y11) of one unit block of luminance data (Y) are one element value (U00) of luminance data (U). -11). That is, the decoded data of each unit block of the color difference data (U, V) is divided into four, and is used separately for the decoded data of the four unit blocks of the luminance data (Y). Accordingly, the processes from Step # 21 to Step # 25 are processed at a frequency of once every four times for the luminance data (Y) for the color difference data (U, V).

  When the format reverse conversion process in step # 27 is completed, it is determined whether the completed unit block is the last unit block (step # 26). If it is not the last unit block, the next unit block is selected ( Step # 27) and the processing from step # 21 to step # 26 are repeated until it is determined as the last unit block.

  Next, the data compression rate from the input image data of the image data encoded by the encoding device 30 of the present invention will be examined.

  Due to the format conversion from the input image data to the primary image data by the format conversion means 31, the data amount is reduced to 50%. For the YUV411 format, when the encoding process is performed under the same conditions (threshold Dth = 13) as in the first embodiment, the data compression rate for the primary image data is about 40%. The data compression rate can be reduced to 20%.

<Third Embodiment>
Next, a third embodiment of the encoding apparatus and method of the present invention will be described with reference to FIG. The encoding apparatus and method of the present invention according to the third embodiment is an embodiment in which the variation degree determination means 13 and the variation degree determination step (step # 13) in the first embodiment and the second embodiment are changed. .

  In the first embodiment and the second embodiment, a standard in which unit blocks having a small degree of variation in element values within 2 × 2 pixels exist in a condition of a predetermined threshold value Dth (for example, 13) is more than half of the whole. Assuming a natural image of a photographic image or a moving image as the primary image data, the storage capacity of the image memory is set. If there are more than the initial predictions, unit blocks are sequentially selected, and the processing from steps # 12 to # 18 shown in FIG. 6 is repeated until it is determined to be the last unit block. In addition, when the storage capacity of the image memory is exhausted, the last several unit blocks are encoded data ECD9 or ECD2 regardless of the degree of variation. It may not be able to store in the image memory is generated.

  Therefore, in the third embodiment, in order to prevent such a storage capacity shortage of the image memory, the variation degree determination means 13 replaces the variation degree determination step (step # 13) shown in FIG. The variation degree determination step (step # 130) shown in FIG. First, it is determined whether or not the cumulative count number M of unit blocks determined to have a large degree of variation is equal to or less than half the total number of unit blocks of RGB or YUV components (3456000 for RGB data) (step # 131). If the cumulative count number M is less than half of the total number of unit blocks in the determination of step # 131, sufficient storage capacity for creating the 12-bit partial encoded data ECD12 of the encoded data ECD21 remains in the image memory. Therefore, the variation degree is determined by the same variation degree determination step (step # 13) as in the first embodiment, and the flag bit FLG is set (step # 132). Next, when the flag bit FLG is set to “1”, the cumulative count number M is incremented by 1 (M = M + 1, step # 133), and the process proceeds to step # 14. If the flag bit FLG is set to “0” in step # 132, the process proceeds to step # 17 without incrementing the cumulative count number M. Note that the initial value of the cumulative count number M is zero. On the other hand, if the cumulative count number M is larger than half of the total number of unit blocks in the determination of step # 131, encoding is performed on the surplus portion after the storage capacity for 9-bit encoded data ECD9 is secured in the image memory. Since sufficient storage capacity for creating the 12-bit partial encoded data ECD12 in the data ECD21 is not left, variation based on the absolute value (| Pij−AVG |) of the difference between each element value Pij and the average value AVG Step # in the case where the flag bit FLG is set to “0” (step # 134) and the flag bit FLG is set to “0” on the assumption that the degree of variation is forcibly small without performing the magnitude judgment. 17 is forcibly transferred.

  Since it is the same as the said 1st Embodiment and 2nd Embodiment except the variation degree determination process, the overlapping description is omitted.

  Furthermore, after the encoded data generation step 1 or 2 is completed for all block units, if the cumulative count number M exceeds half of the total number of unit blocks, the following is performed according to the excess amount. A threshold resetting step is provided for resetting the threshold Dth to a higher value to cope with the primary image data encoding process. Here, the threshold value Dth may be increased by 1 regardless of the excess.

  On the contrary, in the threshold resetting step, when the cumulative count number M is less than half of the total number of unit blocks and the surplus number is greater than or equal to a predetermined specified value, it corresponds to the encoding process of the next primary image data. Processing for resetting the threshold value Dth as high as possible may be added. Here, the threshold value Dth may be lowered by 1 regardless of the surplus number.

<Another embodiment>
<1> In each of the embodiments described above, the deviation value calculation means 15 of the encoding devices 10 and 30 of the present invention calculates the deviation value SD based on the equation (4). You may calculate based on the following formula 7 which calculates a standard deviation.

  <2> In each of the above embodiments, the variation degree determination means 13 of the coding apparatuses 10 and 30 of the present invention determines the variation degree of the distribution of the four element values Pij of each unit block, among the four pixels in the unit block. Is determined to be large if the absolute value (| Pij−AVG |) is greater than or equal to the threshold value Dth, and the absolute value (| When Pij−AVG |) is less than the threshold value Dth, it is determined to be small. However, instead of the determination criterion based on the magnitude comparison between the absolute value (| Pij−AVG |) and the threshold value Dth, the deviation value calculation unit 15 calculates. A criterion for determining that the degree of variation is large when the deviation value SD is greater than or equal to a predetermined threshold and that the degree of variation is small when the deviation value SD is less than the predetermined threshold may be used.

  However, in this case, the deviation value calculating unit 15 needs to calculate the deviation value SD in advance before the variation degree determining unit 13 determines the degree of variation, regardless of the unit block variation degree. 6 and 11, the deviation value calculation step of step # 15 is processed between the average value calculation step of step # 12 and the variation degree determination step of step # 13. Thus, the process order is changed.

  <3> In each of the above embodiments, the number of bits of the deviation value SD is the same as the number of bits of the element value Pij of the unit block, but it may be reduced by 1 bit. That is, in the case where each element value Pij is 8-bit gradation data, the deviation value SD is also 8 bits, but may be reduced by 1 bit to 7 bits. More specifically, when the element value Pij is 256 gradations (gradation values 0 to 255), the average value AVG calculated by Equation 3 can also take gradation values 0 to 255, but the deviation value calculated by Equation 4 is used. SD becomes “128” when the element values Pij (i = 0, 1, j = 0, 1) are 0, 0, 255, and 255, and the degree of variation is maximum. Since the maximum value “128” of the deviation value SD is a result of rounding off “127.5”, if this is set to “127”, the deviation value SD can be represented by 7 bits. Therefore, the 21-bit encoded data ECD21 can be 20-bit encoded data that is one bit fewer. Thereby, the storage capacity of the image memory 17 can be further reduced.

  In this case, 1 bit reduced by the deviation value SD is used as the flag bit FLG, so that the encoded data of the unit block having a large variation degree has the same number of bits as the encoded data by the conventional BTC encoding method. Therefore, if there is at least one unit block with a small degree of variation, the encoding apparatus and method according to the present invention reliably reduces the data compression rate compared to the conventional BTC encoding method, and the amount of data after compression is reduced. Will be reduced.

  <4> In each of the above embodiments, the size of the unit block is 2 × 2 pixels or 2 × 2 data holding units. However, the size may be set to 4 × 4 pixels or 4 × 4 data holding units that are twice the height and width. I do not care. However, if the unit block size is simply increased, the difference between the element values P′ij of the decoded image data from the element values Pij of the primary image data increases, and the degree of degradation of the image quality increases. The average value AVG is replaced with a representative value of element values Pij (i = 0 to 3, j = 0 to 3) calculated based on a predetermined reference, the deviation value SD is replaced with a quantization step, and the comparison result data Cij (i = 0-3, j = 0-3) may be changed so as to be expressed by a gradation value of 2 bits or more obtained by quantizing the difference between the element value Pij and the representative value by the quantization step. Even when the unit block size is increased in this way, when the degree of variation between the element values Pij in the unit block is small, the comparison result data Cij (or gradation value data) and the deviation value are the same as in the above embodiments. By omitting SD (or quantized data), the data compression rate can be reduced and the amount of data after compression can be reduced.

  <5> In the above embodiments, the first decoding unit 23 and the second decoding unit 24 of the decoding device 20 of the present invention output the decoded data P′ij to the image display device, and the decoding device of the present invention. The 40 format reverse conversion means 41 has explained the application example in which the decoded data after YUV format decoded data P′ij is converted back to the RGB format is output to the image display device. Is not applied directly to the image display device, but as shown in FIG. 14, the decoded data is used as the image data of the previous frame, and the encoding devices 10 and 30 of the present invention and the decoding of the present invention are used. Current frame image data (gradation data) output directly to the image display device (for example, the liquid crystal display device 50) without passing through the conversion devices 20 and 40, and one frame stored in the image memory By comparing the image data of the (gradation data), applications for correcting the gradation of the image data of the current frame is also conceivable. In the application example shown in FIG. 14, the output image data (gradation data one frame before) of the decoding apparatuses 20 and 40 of the present invention and the input image data (gradation data of the current frame) of the coding apparatuses 10 and 30 of the present invention. ) Is input to a lookup table 51 that is determined by comparing gradation values, and the gradation data of the current frame is corrected by the output of the lookup table 51. Also in this application example, it is possible to reduce the storage capacity of the image memory for storing the image data of the previous frame compared to the case where data compression is performed by the conventional BTC encoding processing method. In FIG. 14, the detailed circuit configuration inside the liquid crystal display device 50 is well-known in, for example, the above-mentioned Patent Documents 2 to 4, and thus the description is omitted.

  <6> In the above embodiments, the primary image data is assumed to be data of 640 × 480 pixels in which each component in the RGB format or YUV format is 8 bits (256 gradations). The number of gradations and the resolution are not limited to the image data of the above embodiments. The color display format, the number of gradations of each component, and the resolution can be changed as appropriate. Further, the image data is not limited to a color image, and may be a monochrome image.

  <7> In each of the above embodiments, the data holding unit is 1, 2, or 4 pixels. However, it is also possible to use a data holding unit having a different number of pixels. For example, a data holding unit of 16 pixels in which 4 × 4 pixels in length × width can be used. However, as the number of pixels in the data holding unit, that is, the number of pixels constituting the unit block is larger, the compression can be performed to reduce the amount of data after compression, but the image quality is further deteriorated. Is preferably a data holding unit of 1, 2, or 4 pixels.

  <8> In each of the above embodiments, the flag bit indicating the determination result regarding the degree of variation is set to 1 bit. However, a flag bit having a larger number of bits may be used. However, it is preferable to use one flag bit because the amount of data can be suppressed rather than using a flag bit having a larger number of bits.

  INDUSTRIAL APPLICABILITY The present invention can be used for an image encoding apparatus and method for compressing image data, and an image decoding apparatus and method for expanding compressed image data. For example, gradation data for driving a liquid crystal display device is used. The present invention can be applied to the reduction of the storage capacity of an image memory that temporarily stores frame image data in the overdrive method for correction.

1 is a block diagram illustrating an example of a system configuration in an image encoding device and an image decoding device according to a first embodiment of the present invention. The figure which shows the structural example of the RGB format image data of an encoding process target The figure which shows the structural example of the unit block used with the image coding method and image decoding method which concern on this invention The figure which shows typically the data structure of two types of encoding data produced | generated with the image coding apparatus which concerns on this invention. The figure which shows the structural example of the image memory used with the image coding apparatus which concerns on this invention. Process drawing which shows an example of the rough process sequence in 1st Embodiment of the image coding method which concerns on this invention Process drawing which shows an example of the schematic process sequence in 1st Embodiment of the image decoding method which concerns on this invention The figure which shows a specific example of the process result in 1st Embodiment of the image coding method and image decoding method which concern on this invention. The figure which shows typically the data structure of YUV422 format and YUV411 format corresponding to RGB data for 4x4 pixels The block diagram which shows the system configuration example in 2nd Embodiment of the image coding apparatus and image decoding apparatus which concern on this invention Process drawing which shows an example of the schematic process sequence in 2nd Embodiment of the image coding method which concerns on this invention Process drawing which shows an example of the schematic process sequence in 2nd Embodiment of the image coding method which concerns on this invention Process drawing which shows the process sequence of the variation degree determination process in 3rd Embodiment of the image coding method which concerns on this invention. 1 is a block diagram showing an application example of an image encoding device and an image decoding device according to the present invention. Process diagram showing an example of a processing procedure of encoding processing by a conventional BTC encoding method and a diagram showing a configuration example of a pixel block of 2 × 2 pixels The figure which shows the structural example of the encoding data by the conventional BTC encoding system. Process diagram showing an example of processing procedure of decoding process by conventional BTC encoding method and figure showing configuration example of decoded data The figure which shows the specific example of the encoding system by the conventional BTC encoding system, and a decoding process The figure which shows the specific example in case the difference between the encoding method by the conventional BTC encoding method and the pixel value of the pixel block of a decoding process is not remarkable.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10, 30: Image coding apparatus which concerns on this invention 11: Image data division | segmentation means 12: Average value calculation means 13: Dispersion degree determination means 14: Comparison result data generation means 15: Deviation value calculation means 16: Encoding data generation means 17: Image memory 20, 40: Image decoding device according to the present invention 21: Data reading means 21a: First data reading means 21b: Second data reading means 22: Dispersion degree identifying means 23: First decoding means 24: Second decoding means 31: Format conversion means 41: Format reverse conversion means 50: Liquid crystal display device 51: Look-up table

Claims (15)

Image data dividing means for dividing image data composed of pixel data having a predetermined number of pixels as a data retention unit into unit blocks composed of a plurality of adjacent data retention units;
Average value calculating means for calculating an average value of element values of each data holding unit in the unit block generated by the image data dividing means for each unit block;
For each unit block, a variation for generating a flag bit indicating the determination result by determining the degree of variation in the distribution of element values of each data holding unit in the unit block based on a predetermined criterion. Degree determination means;
Comparison result data generation means for generating, for each unit block, comparison result data indicating each magnitude comparison result of the element value of each data holding unit in the unit block and the average value calculated by the average value calculation means When,
For each unit block, the element of each data holding unit in the unit block based on the difference value between the element value of each data holding unit in the unit block and the average value calculated by the average value calculation means Deviation value calculating means for calculating a deviation value indicating an average deviation degree of the value from the average value;
For each unit block, according to the value of the flag bit generated by the variation degree determination unit, if the variation degree is small, the average value and the flag bit are combined to generate encoded data, And an encoded data generation unit configured to generate encoded data by combining the average value, the flag bit, the comparison result data, and the deviation value when the degree of variation is large. Encoding device.   The variation degree determination means determines whether the absolute value of the difference value between the element value of the data holding unit and the average value calculated by the average value calculating means is predetermined for all the data holding units in the unit block. The image coding apparatus according to claim 1, wherein when it is smaller than a threshold value, it is determined that the variation degree is small, and in other cases, it is determined that the variation degree is large.   The variation degree determination means determines that the variation degree is small when the deviation value is smaller than a predetermined determination threshold value, and determines that the variation degree is large when the deviation value is equal to or larger than the predetermined determination threshold value. The image coding apparatus according to claim 1, wherein:   The encoded data generation means generates first encoded data composed of the average value and the flag bit for all the unit blocks regardless of the value of the flag bit, and depends on the value of the flag bit. Whether the unit block has a large degree of variation is identified, and second encoded data composed of the comparison result data and the deviation value is generated only for the unit block having a large degree of variation. The image coding apparatus according to any one of claims 1 to 3, wherein   An image memory for storing the encoded data is provided, and a memory space of the image memory is divided into a first area for storing the first encoded data and a second area for storing the second encoded data. The image coding apparatus according to claim 4, wherein the image coding apparatus is configured. In the process in which the encoded data generation unit sequentially generates the encoded data for each unit block, and the accumulated data amount of the encoded data exceeds a predetermined upper limit value, the encoded data, And the encoded data to be generated thereafter is generated without combining the comparison result data and the deviation value,
The image coding apparatus according to claim 1, wherein the determination criterion used by the variation degree determination unit is changed so that the variation degree is easily determined to be small.   When the pixel data is composed of gradation data of a plurality of different color components in one pixel, one color component in one pixel is treated as the data holding unit, the image data dividing means, the average value calculation The means, the variation degree determination means, the comparison result data generation means, the deviation value calculation means, and the encoded data generation means each process differently for each color component. The image encoding device according to any one of the above.   When the pixel data is composed of luminance data in units of pixels and two types of color difference data in units of pixels or two or four pixels adjacent to each other, for the luminance data, one pixel is stored as the data For the two types of color difference data, 1, 2 or 4 pixels are treated as the data holding unit, the image data dividing means, the average value calculating means, the variation degree determining means, and the comparison result data generating The means, the deviation value calculation means, and the encoded data generation means each process the data separately for the luminance data and the two types of color difference data. The image encoding device according to item 1.   Original image data in which the pixel data is composed of gradation data for each color component in pixel units, luminance data in pixel units, and pixel units or two adjacent pixels or two in four pixel units. 9. The image coding apparatus according to claim 8, further comprising image format conversion means for converting into image data composed of different types of color difference data. An image decoding device that decodes the encoded data encoded by the image encoding device according to any one of claims 1 to 6 into image data including pixel data of the data holding unit,
Data reading means for individually reading out the average value and the flag bit constituting the encoded data or the average value, the flag bit, the comparison result data, and the deviation value for each unit block;
Based on the value of the flag bit read by the data reading means, a variation degree identifying means for identifying the degree of variation in the distribution of element values of each data holding unit in the unit block;
For the unit block identified as having a large variation degree by the variation degree identifying unit, decoded data is generated based on the average value, the comparison result data, and the deviation value read by the data reading unit. First decoding means;
Second decoding means for generating decoded data based on the average value read by the data reading means for the unit block identified as having a small degree of variation by the variation degree identifying means. An image decoding apparatus characterized by comprising: The first decoding means calculates the element value of each data holding unit after decoding in the unit block based on the comparison result data corresponding to each data holding unit with respect to the average value. Calculate by adding or subtracting the deviation value,
The image decoding apparatus according to claim 10, wherein the second decoding unit sets all the element values of the data holding units after decoding in the unit block as the average value.   The data reading means is configured to read from the encoded data when the variation degree is identified by the first data reading means for reading the average value and the flag bit from the encoded data and the variation degree identifying means. The image decoding apparatus according to claim 10 or 11, further comprising second data reading means for reading the comparison result data and the deviation value. The image encoding device according to any one of claims 1 to 4,
An image memory for storing encoded data encoded by the image encoding device;
An image data processing apparatus comprising: the image decoding apparatus according to any one of claims 10 to 12 that decodes encoded data stored in the image memory. An image data dividing step of dividing image data consisting of pixel data having a predetermined number of pixels as a data holding unit into unit blocks consisting of a plurality of adjacent data holding units;
For each unit block, an average value calculating step for calculating an average value of element values of each data holding unit in the unit block generated in the image data dividing step;
For each unit block, a variation for generating a flag bit indicating the determination result by determining the degree of variation in the distribution of element values of each data holding unit in the unit block based on a predetermined criterion. Degree determination step,
A comparison result data generating step for generating, for each unit block, comparison result data indicating each size comparison result of the element value of each data holding unit in the unit block and the average value calculated in the average value calculating step. When,
For each unit block, the element of each data holding unit in the unit block based on the difference value between the element value of each data holding unit in the unit block and the average value calculated in the average value calculating step A deviation value calculating step of calculating a deviation value indicating an average deviation degree of the value from the average value;
For each unit block, according to the value of the flag bit generated in the variation degree determination step, if the variation degree is small, the average value and the flag bit are combined to generate encoded data, An encoded data generating step of generating encoded data by combining the average value, the flag bit, the comparison result data, and the deviation value when the degree of variation is large; Encoding method. An image decoding method for decoding the encoded data encoded by the image encoding method according to claim 14 into a data format before encoding,
A first data reading step for individually reading out the average value and the flag bit constituting the encoded data for each unit block;
Based on the value of the flag bit read in the data reading step, a variation degree identifying step for identifying the degree of variation in the distribution of element values of each data holding unit in the unit block;
A data second reading step for individually reading out the comparison result data and the deviation value constituting the encoded data when the variation degree is identified as being large in the variation degree identifying step;
When it is identified that the degree of variation is large in the variation degree identifying step, the average value read in the first data reading step, the comparison result data read in the second data reading step, and the deviation value A first decryption step for generating decrypted data based on;
And a second decoding step for generating decoded data based on the average value read in the first data reading step when the variation degree is identified as being small in the variation degree identifying step. An image decoding method characterized by the above.

Images