JP6081869B2 - Image decoding device - Google Patents

Description

  The present invention relates to an image decoding apparatus that decompresses and decodes encoded image data.

  2. Description of the Related Art Conventionally, there is an image encoding device that compresses and encodes image data for the purpose of reducing the burden on communication, etc., and decompresses the encoded image data that is the inverse process, and restores the original image Decoding devices are also known. In recent years, an encoding standard that can be compressed at a higher compression rate has been proposed (see Non-Patent Document 1).

  FIG. 9 shows a schematic configuration of an image decoding apparatus according to the standard. As shown in FIG. 9, the image decoding apparatus 100 includes a bit stream analysis unit 110, a VLC decoding unit 120, an adaptive scanning unit 130, a DC / LP / HP inverse prediction processing unit 140, and an inverse The quantization unit 150 includes a frequency inverse transform unit 160 and a color space inverse transform unit 170.

  The bitstream analysis unit 110 analyzes the input bitstream, extracts image information, separates encoded image data, and outputs the separated image data to the VLC decoding unit 120.

  The encoded image data includes 16 × 16 pixels (256 pixels) as one macroblock among pixels (pixels) which are the minimum unit of an image, and the macroblock is 4 × 4 pixels (16 pixels). From the highest frequency band divided into blocks and frequency-converted during compression, the first layer data (16 blocks HP coefficients), the second layer data (one block LP coefficients), and the third layer data (one pixel) Of the DC coefficient).

  For the above-mentioned one block (4 × 4 pixels), the first layer data is blank in the upper left of 4 × 4 pixels, the remaining 15 pixels are HP (high pass) coefficients, and the second layer data is 4 × The upper left of the four pixels is blank, the remaining 15 pixels are LP (low pass) coefficients, and the third layer data is composed of one pixel DC (direct current) coefficient which is a direct current component.

  The VLC decoding unit 120 performs VLC (Variable Length Coding) decoding processing on the first layer data, the second layer data, and the third layer data that are the image encoded data separated by the bit stream analysis unit 110. The non-zero coefficient data and the zero run length data are extracted and output to the adaptive scan unit 130.

  The adaptive scan unit 130 replaces the coefficient data for each block in the reverse order of the compression for each of the first layer data and the second layer data whose pixel order is replaced by a method determined at the time of compression. The process of returning to the pixel order is performed. This replacement process is performed by collecting non-zero coefficients on the lower left side of the frequency band at the top left of the block during compression (conversely, those having zero coefficient values are gathered on the higher side of the lower right frequency band). In order to increase the compression ratio.

  As shown in FIG. 10, the adaptive scan unit 130 includes a non-zero coefficient storage / coefficient total number accumulation unit 131, a coefficient data replacement processing unit 132, a coefficient data storage unit 133, a replacement information storage unit 134, and a frequency update processing unit 135. And a non-zero coefficient storage / coefficient total accumulation unit 131, as shown in FIG. 11, a non-zero coefficient total accumulation unit 141, a coefficient output control unit 142, a non-zero coefficient storage unit The coefficient data replacement processing unit 132 includes an output unit 143 and a zero run length storage / output unit 144. The coefficient data replacement processing unit 132 includes a coefficient data holding unit 145, a replacement processing control unit 146, a coefficient position coordinate update unit 147, a coefficient, as shown in FIG. An address holding unit 148 and a block completion signal generation unit 149 are included.

  Hereinafter, processing in the adaptive scanning unit 130 will be described. In the following, the processing of the first layer data will be described, but the same applies to the processing of the second layer data. 13 to 22 are explanatory diagrams for explaining the adaptive scan processing.

  The first layer data input from the VLC decoding unit 120 is handled in units of blocks corresponding to 4 × 4 pixels as shown in FIGS. 20 represents the position (coordinate position) of the pixel in the block, and the HP coefficient shown in FIG. 21 is shown in FIG. 16 in the data table format.

In the replacement process, the non-zero coefficient data C (i) corresponding to the non-zero coefficient forward coordinate i which is the arrangement order in the VLC code of the non-zero coefficient data, which is variable-length decoded by the VLC decoding unit 120, and the The non-zero coefficient data C (i) is included in the zero run length data R (i) indicating the number of zero coefficients preceding the non-zero coefficient C (i), and the immediately preceding non-zero coefficient data includes the zero coefficient. J _PRE + R (i) +1 obtained by adding the zero run length R (i) and 1 to the coefficient position coordinate j _PRE (the initial value for the first non-zero coefficient in the block is 0) that is the coefficient position of the new coefficient position As the coefficient data Z (P (j)) with the coordinate j, the coefficient positions are switched and output. The replaced Z (P (j)) is represented as Z ′ (j). Further, although not generated as intermediate data, coefficient data when a zero coefficient before coefficient replacement is included is represented as Z (j). The frequency update process for updating the non-zero frequency data after the coefficient replacement process reads the non-zero frequency data F (j) stored in the replacement information storage unit 134 and reads the read non-zero frequency data F (j ) Is updated by adding 1, and the updated non-zero frequency data is F ′ (j). Then, the comparison / replacement process for replacing the replacement data P (j) and the frequency update data F ′ (j) is the coefficient data of the coefficient position coordinate j to which the non-zero coefficient data C (i) is to be transferred. And non-zero frequency data F ′ (j) representing the relationship with the frequency where Z (j) was non-zero, and replacement data P (j) for replacing coefficient data Z (j) in the block It is done using. The replacement data P (j) is set in advance according to the prediction mode in the DC / LP / HP inverse prediction processing unit 140, and the prediction data derived in the DC / LP / HP inverse prediction processing unit 140 is used. Corresponding replacement data P (j) is used. Further, the replacement data P (j) and the non-zero frequency data F (j) are set in advance with initial values. An example of initial values of the replacement data P (j) and the non-zero frequency data F (j) is shown in FIG. These replacement data P (j) and non-zero frequency data F (j) are stored in the replacement information storage unit 134.

  In FIG. 10, the DC / LP / HP inverse prediction processing unit 140 stores information on the prediction mode, and the VLC decoding unit 120 stores the first layer data for each block of non-zero coefficient storage / coefficient total number accumulation unit 131. Is input. The non-zero coefficient storage / coefficient total accumulation unit 131 first stores non-zero coefficient data C (i) and zero run length data R (i), which are first layer data (HP coefficient data) for one block. The total number m of non-zero coefficient data for one block is accumulated. After the accumulation is completed, the non-zero coefficient data C (i) and the zero run length data R (i) are output to the coefficient data replacement processing unit 132 one coefficient at a time. The non-zero coefficient storage / coefficient total accumulation unit 131 processes the non-zero coefficient data C (i) and zero run length data R (i), which are data of the next coefficient, for a total number m of non-zero coefficients. Is output to the coefficient data replacement processing unit 132 every time a single coefficient replacement completion signal indicating that replacement (update) of replacement information data has been completed is received.

  Specifically, in FIG. 11, the non-zero coefficient data C (i) input from the VLC decoding unit 120 is output to and stored in the non-zero coefficient storage / output unit 143, and zero run length data R (i) Is output to and stored in the zero run length storage / output unit 144, and 1 is added to the non-zero coefficient total number m in the non-zero coefficient total number accumulating unit 141 to obtain a new non-zero coefficient total number m. The process is repeated until a block decoding completion signal is received from the VLC decoding unit 120 indicating that the coefficient is a non-zero coefficient. Upon receiving the block decoding completion signal, the non-zero coefficient total accumulation unit 141 outputs the accumulated non-zero coefficient total number m to the coefficient output control unit 142 and indicates that the total of the non-zero total has been completed. A completion signal is output to the coefficient output control unit 142. The coefficient output control unit 142 that has received the total accumulation completion signal outputs a coefficient output command signal to the non-zero coefficient storage / output unit 143 and zero run length storage / output unit 144, and receives the coefficient output command signal. The zero coefficient storage / output unit 143 and the zero run length storage / output unit 144 output the non-zero coefficient data C (i) and the zero run length data R (i) in the block to the coefficient data replacement processing unit 132 by one coefficient at a time. . The coefficient output control unit 142 outputs a block output completion signal to the coefficient data replacement processing unit 132 after outputting the m-th non-zero coefficient data C (m) and zero run length data R (m).

  The coefficient data replacement processing unit 132 to which the non-zero coefficient data C (i) and the zero run length data R (i) are input outputs a new coefficient position coordinate j to the replacement information storage unit 134 and the replacement information. The value of the replacement data P (j) read from the storage unit 134 (which is an initial value in the case of the first block) is used as an input address, and the non-zero coefficient data C (i) is used as input data. Output to.

Specifically, in FIG. 12, the coefficient data replacement processing unit 132, to which the non-zero coefficient data C (i) and the zero run length data R (i) are input, converts the non-zero coefficient data C (i) into coefficient data. The data is output as input data to be stored in the coefficient data storage unit 133 through the holding unit 145, and the zero run length data R (i) is output to the coefficient position coordinate update unit 147 through the replacement processing control unit 146. the coefficient coordinates updating unit 147, the non-zero coefficient data immediately before is a coefficient position when containing zero coefficients coefficients coordinates j _PRE, j _{PRE +} R plus the zero run length and 1 R (i) (I) +1 is output to the replacement information storage unit 134 as a new coefficient position coordinate j. Note that the updated j is assigned to j _PRE to prepare for the next non-zero coefficient processing. The replacement information storage unit 134 outputs the replacement data P (j) of the new coefficient position coordinate j to the coefficient data storage unit 133 via the coefficient address holding unit 148 as a data address. The non-zero coefficient storage / coefficient total accumulation unit 131 outputs the m-th non-zero coefficient data C (m) and zero run length data R (m), which are the total number of non-zero coefficients, and then stores the non-zero coefficient storage / coefficients. The block completion signal generation unit 149 to which the block output completion signal is input from the total number accumulation unit 131 outputs the block replacement completion signal to the coefficient data storage unit 133 after the single coefficient replacement completion signal is output from the replacement processing control unit 146. Output.

  The coefficient data storage unit 133 stores the non-zero coefficient data C (i) as data in coefficient data Z (P (j)) with replacement data P (j) as an address position. The replaced Z (P (j)) is represented as Z '(j), and is not generated as intermediate data, but the coefficient data including the zero coefficient before the replacement is represented as Z (j). The coefficient data storage unit 133 that has received the block replacement completion signal from the coefficient data replacement processing unit 132 receives the replaced coefficient data Z ′ (1) to Z ′ (15) at a time in the DC / LP / HP reverse. The result is output to the prediction processing unit 140. The pre-replacement coefficient data Z (j) column of FIG. 18 shows the data before the replacement process, and the post-replacement coefficient data Z ′ (j) column of FIG. 18 shows the data after the replacement process. ing. In the example shown in FIG. 18, according to the replacement data P (j), the coefficient data Z (4) (value is 9A) at the coordinate position 4 before replacement processing is Z ′ (8) at the coordinate position 8 of the new block. Similarly, the coefficient data Z (5) (value 11) at the coordinate position 5 before the replacement processing is incorporated into Z ′ (2) at the coordinate position 2 of the new block. The first hierarchy data generated in this way is shown in FIG.

  Next, after the replacement of one Z ′ (j) is completed, the coefficient data replacement processing unit 132 outputs a frequency update completion signal for starting the frequency update processing to the frequency update processing unit 135 and updates The new coefficient position coordinate j is output to the replacement information storage unit 134, and then the replacement information storage unit 134 performs the frequency update process on the non-zero frequency data F (j) corresponding to the new coefficient position coordinate j. To the unit 135.

  The frequency update processing unit 135 adds 1 to the non-zero frequency data F (j) input from the replacement information storage unit 134, and the comparison / replacement processing unit 136 and the replacement are obtained as non-zero frequency data F ′ (j). The information is output to the information storage unit 134. The replacement information storage unit 134 stores the non-zero frequency data F ′ (j) and updates it as F (j) for processing of the next block.

  FIG. 18 shows the table data after this addition processing. In the example shown in FIG. 18, since the coefficient data Z (4) at the coordinate position 4 is 9A and the coefficient data Z (5) at the pixel position 5 is 11 and non-zero, the non-zero frequency data F ( 4) 1 is added to = 28, and the updated non-zero frequency data F ′ (4) becomes 29, and 1 is added to non-zero frequency data F (5) = 24 at the coordinate position 5 to update the non-updated non-frequency data F ′ (4). The zero frequency data F ′ (5) is 25. The frequency update processing unit 135 outputs the updated non-zero frequency data F ′ (j) thus added to the comparison / replacement processing unit 136 for each non-zero coefficient.

  The comparison / replacement processing unit 136 compares the non-zero frequency data F ′ (j−1) read from the replacement information storage unit 134 with F ′ (j) input from the frequency update processing unit 135, and F When '(j) is larger than F' (j-1), F '(j-1) and F' (j) are replaced, and F "(j-1) = F '(j) , F ″ (j) = F ′ (j−1), and replacement data P (j−1) and P (j) before and after read from the replacement information storage unit 134 are replaced by P Processing is performed such that '(j-1) = P (j) and P' (j) = P (j-1) '.

  FIG. 19 shows the update data after this replacement processing. The example shown in FIG. 19 is obtained by replacing the one shown in FIG. 18, and F ′ (4)> F ′ (3) in the state before the replacement shown in FIG. P (3) at position 3 and P (4) at coordinate position 4 are replaced with P ′ (3) = 8, P ′ (4) = 5, and F ′ (3) and coordinate at coordinate position 3 F ′ (4) at position 4 is replaced with F ″ (3) = 29 and F ″ (4) = 28.

  Note that the table data F (j) and P (j) stored in the replacement information storage unit 134 are F ″ (j−1) and F ″ (j) output from the comparison / replacement processing unit 136, respectively. And P ′ (j−1) and P ′ (j).

  In the adaptive scanning unit 130, the above-described processing is performed, and first hierarchical data in which coefficient data is replaced according to non-zero frequency and preset replacement data is generated. Similar processing is performed on the second layer data, and second layer data in which the coefficient data is replaced is generated. The third layer data is output to the DC / LP / HP inverse prediction processing unit 140 as it is.

  FIG. 13 is a flowchart showing the operation of the adaptive scan unit 130. At the beginning of the block processing, after initializing by substituting 0 into Z (1) to Z (15) in S101, In the VLC decoding process of S103, the process of obtaining non-zero coefficient data in C (i) and zero-run length data in R (i) by variable length decoding is repeated for the number of non-zero coefficients in the block. Using the acquired non-zero coefficient data C (i) and zero run length data R (i), the replacement process to the coefficient data Z ′ (j) is performed in S107, and the non-zero data F (j) is updated in S108. Thereafter, based on the comparison of the updated non-zero data F ′ (j) and F ′ (j−1), substitution processing of F ′ (j) and P (j) is performed in S110 to S112, and S107 to S112. The non-zero handler obtained in step S106 It is intended to repeat the total number of m times.

  FIG. 14 shows the operation of the adaptive scanning unit 130 as a state transition diagram. After the processing starts, the state 0 of ST0 that is in a standby state transitions to ST1 for processing of the next block, and ST1 In state 1, after acquiring non-zero coefficients and zero run length by the number of non-zero coefficients of the block by VLC decoding, the state transitions to state 2 of ST2. In state 2, coefficient data replacement processing, frequency update processing, and comparison / replacement processing are performed for each coefficient. When processing for the number of non-zeros is not performed, the state transitions again to state 2 and non-zero. The processing in state 2 is repeated until the processing for the number of items is completed. When the replacement processing, which is the last processing in state 2, is completed, if processing for the number of non-zeros has been completed, the state transits to ST0, which is waiting, and waits for processing of the next block. .

  FIG. 15 is a timing chart showing the operation of the adaptive scanning unit 130. This timing chart is an example in which the number of non-zero coefficients is three, and the processing time of each process varies depending on the circuit configuration. In FIG. 15, since the coefficient replacement process, the frequency update process, and the comparison / replacement process are performed serially after the VLC decoding processes from vlc1 to vlc3 are performed together, many cycles are consumed for the process. In the embodiment of FIG. 15, the processing time for one block is 33 cycles.

  The DC / LP / HP inverse prediction processing unit 140 receives the first layer data, the second layer data, and the third layer data for each block, which are input from the adaptive scan unit 130, between adjacent blocks. The prediction mode is derived by comparing the coefficients at, and the value of the corresponding pixel of the highly correlated block is added to the target pixel according to the set arithmetic expression with respect to the prediction mode, so that the prediction process before compression is performed. Is restored (inverse prediction processing). Then, the processed first layer data, second layer data, and third layer data are output to the inverse quantization unit 150 for each block.

  The inverse quantization unit 150 multiplies the first layer data, the second layer data, and the third layer data, which have been subjected to the inverse prediction processing by the DC / LP / HP inverse prediction processing unit 140, with respective set quantization parameters. Then, the inverse quantization process is performed, and the processed first layer data, second layer data, and third layer data are output to the frequency inverse transform unit 160 for each block.

  The frequency inverse transform unit 160 sequentially processes the image data inversely quantized by the dequantization unit 150 in a raster direction with 4 × 4 pixels as one block, and generates coefficient data for each frequency band as pixel data. Process to convert to. The frequency inverse transform unit 160 converts the first layer data, the second layer data, and the third layer data separated for each frequency band into color components (Y / U / V or Y / U / V / K). Each time, the data is returned to one block (4 × 4 pixels) of two-dimensional space data and output to the color space inverse transform unit 170.

  The color space inverse transform unit 170 processes the image data for each color component input from the frequency inverse transform unit 160 for each pixel, and when the color space inverse transform is not necessary, the color space inverse transform is not performed. When reverse conversion is necessary, if the color component is Y / U / V, it is converted to RGB data, and if the color component is Y / U / V / K, it is converted to CMYK data, and the processed pixel The data is output to the outside as image data for one image.

JPEG XR Recommendation ITU-T T.832: Information technology -JPEG XR image coding system-Image coding specification. [Online] .Recommendation T.832 (01/12). Approval 2012-01-13. Status: In force. Retrieved from the Internet: <URL: http://www.itu.int/rec/T-REC-T.832-201201-I/en>.

  By the way, the amount of data of image data to be compressed has increased rapidly due to the improvement in resolution accompanying technological advancement. In the conventional image decoding apparatus 100, the processing takes a long time due to the increase in data amount. Therefore, further efficiency is required.

  In particular, the adaptive scan unit 130 stores the non-zero coefficient C (i) and the zero-run length R (i) for one block, and then collects the total number of non-zero coefficient data in the block. The replacement process to the coefficient data Z ′ (j) is performed, the non-zero frequency data F (j) is updated, and the updated non-zero data F ′ (j) and replacement data P (j) are replaced. As a result, the efficiency of the process cannot be improved, and the process takes a long time. For this reason, the processing of the conventional adaptive scanning unit 130 has become one of the bottleneck factors that reduce the processing efficiency of the image decoding apparatus 100.

  The present invention has been made in view of the above circumstances, and it is an object of the present invention to provide an image decoding apparatus capable of increasing the overall processing efficiency by increasing the processing efficiency of the adaptive scanning unit. To do.

In order to solve the above problems, an image decoding apparatus according to the present invention provides:
A decoding unit that decodes the compressed image encoded data to generate non-zero coefficient data C (i) and zero-run length data R (i) of each non-zero coefficient forward coordinate i in block units;
The coefficient position coordinate j of the block generated from the zero run length data R (i) and the coefficient data of the coefficient position coordinate j to which the non-zero coefficient data C (i) is to be transferred, Z (j) is non- Non-zero frequency data F (j) representing the relationship with the frequency that was zero, and replacement data P (j) for replacing the coefficient data Z (j) in the block, the current coefficient position coordinate j And the block decoded by the decoding unit using the replacement data P (j) representing the relationship between the coefficient position coordinate j ′ of the coefficient position coordinate j and the coefficient position coordinate j ′ to be incorporated by replacement. An adaptive scan unit that replaces the unit non-zero coefficient data C (i) in the block;
An inverse quantization unit that inversely quantizes the n coefficient data in units of blocks replaced by the adaptive scan unit;
In an image decoding apparatus comprising at least a frequency inverse transform unit that inversely frequency transforms the inversely quantized coefficient data in units of a predetermined number of pixels to generate image data,
The adaptive scan unit includes:
A replacement information storage unit for storing the non-zero frequency data F (j) and replacement data P (j);
Coefficient data Z (j) corresponding to the non-zero coefficient data C (i) and zero run length data R (i) decoded by the decoding unit is replaced with replacement data P (j stored in the replacement information storage unit. ) Based on the coefficient data replacement processing unit for replacement processing,
Based on the replacement processing information of the replacement data P (j) and the non-zero coefficient data C (i), the non-zero coefficient data C (i) is stored in Z (P (j)) in the block. A coefficient data storage unit that performs replacement processing and outputs it as Z ′ (j);
The non-zero frequency data F (j) is updated by adding 1 to the non-zero coefficient position coordinate j of the non-zero coefficient position coordinate j after the decoding, and the updated non-zero frequency data F ′ (j ) Frequency update processing unit,
Using the non-zero frequency data F ′ (j) updated by the frequency update processing unit, F ′ (j−1) and F ′ (j) are sequentially compared, and F ′ ( j −1) and Based on the comparison result of F ′ ( j ), a comparison / replacement processing unit for performing replacement processing of F ′ ( j− 1) and F ′ ( j ), P ( j− 1) and P ( j ), A coefficient data replacement processing unit configured to output a single coefficient decoding start signal for starting decoding of the next non-zero coefficient in parallel with the replacement processing from the coefficient data replacement processing unit to the decoding unit; Relates to an image decoding apparatus that outputs a frequency update start signal after coefficient data replacement processing from the frequency update processing unit to the frequency update processing unit. However, n is an integer of 2 or more, i is an integer of 1 to n or less, and j and P (j) are integers of 1 to n.

Furthermore, in the image decoding apparatus of the present invention,
The adaptive scan unit outputs the single coefficient decoding start signal for starting decoding of the next non-zero coefficient simultaneously with the start of the replacement process.

And in the image decoding apparatus of the present invention,
The coefficient data replacement processing unit
A coefficient data holding unit that outputs the non-zero coefficient data C (i) input from the decoding unit to the coefficient data storage unit as input data to the storage unit;
A replacement process control unit for controlling the start of the frequency update process and the start timing of the decoding process;
A coefficient position coordinate update unit for generating a current coefficient position coordinate j;
A coefficient address holding unit that obtains replacement data P (j) from the replacement information storage unit using the current coefficient position coordinate j, and outputs it to the coefficient data storage unit as an input address to the coefficient data storage unit;
A block that outputs a block replacement completion signal indicating completion of block replacement to the coefficient data storage unit based on a block decoding completion signal input from the decoding unit and a single coefficient replacement completion signal input from the replacement processing control unit A completion signal generator,
The replacement processing control unit outputs a single coefficient decoding start signal for starting decoding of the next non-zero coefficient to the decoding unit, and outputs a frequency update start signal to the frequency update processing unit after the coefficient data replacement processing. Output.

  According to the image decoding apparatus according to the present invention having the above configuration, first, encoded image data for one image is input to the decoding unit in units of a block having a predetermined number of pixels. In the decoding unit, the encoded image data is converted into non-zero coefficient data C (i) and zero-run length data R (i) on a block basis and input to the adaptive scan unit.

In the adaptive scanning section, each non-zero coefficient forward coordinate i of the block, the coefficient position coordinate j of the block generated from the zero-run length data, and the coefficient position coordinate to which the non-zero coefficient data C (i) is to be transferred non-zero frequency data F (j) representing the relationship with the frequency where Z (j) is non-zero, which is coefficient data of j, and replacement data P for replacing the coefficient data Z (j) in the block (J) using replacement data P (j) representing the relationship between the current coefficient position coordinate j and the coefficient data Z (j) of the coefficient position coordinate j to be incorporated by replacement. The block unit non-zero coefficient data C (i) decoded by the decoding unit is replaced in the block, and the block unit coefficient data Z ′ (j) after replacement is, for example, Reverse After being measured processed, is output to the inverse quantization unit. After the block-unit coefficient data replacement process, the adaptive scanning unit adds 1 to the non-zero coefficient position coordinate j of the non-zero coefficient position coordinate j to update it as non-zero frequency data F ′ (j), Based on the comparison result between the updated F ′ (j−1) and F ′ (j), F ′ ( j− 1) and F ′ ( j ), and P ( j− 1) and P ( j ) The non-zero frequency data F (j) and the replacement data P (j) are updated for the next coefficient processing. In the present invention, at the start of the replacement process in the adaptive scanning unit, a single coefficient decoding start signal for starting decoding of the next non-zero coefficient is output to the decoding unit, so that the decoding unit Decoding processing and a series of coefficient replacement processing, non-zero frequency data update processing, and coefficient comparison / replacement processing in the adaptive scanning unit can be performed in parallel.

  As in the prior art, after decoding for one block is completed in the decoding unit, a series of coefficient replacement processing, non-zero frequency data update processing, and coefficient comparison / replacement processing in the adaptive scan are performed. As a result, the number of cycles required for the processing in the decoding unit and the processing in the adaptive scanning unit cannot be reduced. However, as in the present invention, the processing in the decoding unit and the processing in the adaptive scanning unit in units of blocks. And the processing in the decoding unit and the processing in the adaptive scanning unit are performed in parallel for each coefficient, so that the processing efficiency is improved, and the processing in the decoding unit and the entire processing in the adaptive scanning unit are performed. The number of cycles can be reduced. Note that the number of cycles indicates the number of cycles required to execute the processing, assuming that one clock used as the processing timing is one cycle.

  Next, after the coefficient replacement process by the adaptive scan unit, the block-unit coefficient data Z ′ (j) after the replacement process is subjected to an inverse prediction process and the like, and is dequantized in the inverse quantization unit The frequency inverse transform unit restores the image data by performing frequency inverse transform in units of a predetermined number of pixels.

  According to the image decoding apparatus according to the present invention having the above-described configuration, when the replacement process starts, a single coefficient decoding start signal for starting decoding of the next non-zero coefficient is output to the decoding unit. The processing in the decoding unit and the processing in the adaptive scanning unit can be performed in parallel, and the number of cycles for the entire processing in the decoding unit and the adaptive scanning unit can be reduced.

In the image decoding apparatus of the present invention,
The coefficient data replacement processing unit
The comparison / replacement processing unit receives a single coefficient replacement completion signal indicating that the replacement process has been completed.

And in the image decoding apparatus of the present invention,
The adaptive scanning unit outputs a single coefficient replacement completion signal indicating that the comparison / replacement processing is completed from the comparison / replacement processing unit to the coefficient data replacement processing unit.

  Usually, the decoding processing in the decoding unit is more complicated than the series of coefficient replacement processing, non-zero frequency data update processing, and coefficient comparison / replacement processing in the adaptive scan unit. The process requires more cycles than the process in the adaptive scan unit. Therefore, the decoding process of the next coefficient can be started without confirming the end of the process in the adaptive scan unit, but the number of cycles required for the decoding process is larger than the number of process cycles in the adaptive scan unit. When the number of cycles required for the decoding process is small or the number of cycles is less than the number of processing cycles in the adaptive scanning unit, the processing in the decoding unit and the processing in the adaptive scanning unit cannot be performed normally. . In view of this situation, the present invention further provides a coefficient data replacement process by outputting a single coefficient replacement completion signal indicating completion of update of coefficient data replacement information from the comparison / replacement unit to the coefficient data replacement processing unit. Even if the number of cycles required for the decryption process is small, the unit can wait for the end of the process in the comparison / replacement process unit and instruct the start of the process in the decryption unit. Even when the time is shorter than the processing time in the adaptive scanning unit, the decoding process can be performed normally.

  According to the image decoding apparatus according to the present invention having the above-described configuration, a series of coefficient replacement processing, non-zero frequency data update processing, and coefficient comparison / replacement processing in the adaptive scanning unit are performed in the decoding processing in the decoding unit. Therefore, it is possible to reduce the number of cycles required for the processing in the decoding unit and the entire processing in the adaptive scanning unit.

  Further, according to the adaptive scan unit of the present invention, a signal for starting the decoding process can be output after waiting for the completion of the process in the comparison / replacement processing unit, so that the processing time in the decoding process is adaptive scan. Even if the processing time is shorter than the processing time in the unit, the decoding process can be normally performed.

  As described above, according to the present invention, the process of the adaptive scan unit can be performed in parallel with the decoding process of the decoding unit, and the process of the adaptive scan unit can be performed efficiently. As a result, the overall processing speed of the image decoding apparatus can be increased compared to the conventional case, and an efficient decoding process can be realized. In addition, even when the processing time in the decoding process is shorter than the processing time in the adaptive scan unit, a signal for starting the decoding process can be output after waiting for the comparison / replacement processing unit to end. The decoding process can be normally performed.

It is the block diagram which showed schematic structure of the image decoding apparatus which concerns on one Embodiment of this invention. It is the block diagram of the 1st Example which showed the structure of the adaptive type scan part which concerns on this embodiment. It is the block diagram of the 2nd Example which showed the structure of the adaptive type scan part which concerns on this embodiment. It is the block diagram which showed the structure of the coefficient data replacement process part which concerns on this embodiment. It is the flowchart which showed the process sequence in the adaptive type scan part which concerns on this embodiment. It is the flowchart which showed the process sequence of the coefficient replacement | exchange process in the adaptive type scan part which concerns on this embodiment. It is a state transition diagram showing a processing state in the adaptive scan unit according to the present embodiment. It is an example of the timing chart which showed the processing operation in the adaptive type scan part concerning this embodiment. It is the block diagram which showed schematic structure of the conventional image decoding apparatus. It is the block diagram which showed the structure of the conventional adaptive type scanning part. It is the block diagram which showed the structure of the conventional non-zero coefficient memory | storage and coefficient total number accumulation part. It is the block diagram which showed the structure of the conventional coefficient data replacement process part. It is the flowchart which showed the process sequence in the conventional adaptive type scanning part. It is a state transition diagram showing a processing state in a conventional adaptive scan unit. It is an example of the timing chart which showed the processing operation in the conventional adaptive type scanning part. It is explanatory drawing for demonstrating an example of the HP coefficient within the block of the conventional adaptive type scanning process. It is explanatory drawing for demonstrating an example of the initial value of the replacement data and the non-zero frequency data of the conventional adaptive scanning process. It is explanatory drawing for demonstrating replacement | exchange of the coefficient data and update of non-zero frequency data of the conventional adaptive type scanning process. It is explanatory drawing for demonstrating the replacement data and the non-zero frequency data after the replacement of the conventional adaptive scanning process. It is explanatory drawing for demonstrating the pixel position within the block of the conventional adaptive scanning process. It is explanatory drawing for demonstrating an example which made the HP coefficient of the conventional adaptive type scanning process into the table data format. It is explanatory drawing for demonstrating an example of the HP coefficient after replacement within the block of the conventional adaptive scanning process.

  Hereinafter, specific embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of the image decoding apparatus 1 according to this embodiment.

  As shown in FIG. 1, the image decoding apparatus 1 of the present example includes an adaptive scanning unit 10 having a configuration different from that of the adaptive scanning unit 130 in the conventional image decoding apparatus 100 of FIG. Only in this respect, the configuration of the conventional image decoding apparatus 100 is different. Therefore, the other components are denoted by the same reference numerals as those in the image decoding device 100, and detailed description thereof is omitted.

{First embodiment}
As shown in FIG. 2, the adaptive scan unit 10 according to the first embodiment includes a coefficient data replacement processing unit 20, a coefficient data storage unit 30, a replacement information storage unit 40, a frequency update processing unit 50, and a comparison. As shown in FIG. 4, the coefficient data replacement processing unit 20 includes a replacement processing unit 60. The coefficient data replacement processing unit 20 includes a coefficient data holding unit 21, a replacement processing control unit 22, a coefficient position coordinate update unit 23, a coefficient address holding unit 24, and a block. The completion signal generation unit 25 is configured.

  In the adaptive scan unit 10 as well, as with the adaptive scan unit 150 described above, the first layer data and the second layer data output from the VLC decoding unit 120 are as shown in FIG. 20 and FIG. Processing is performed in units of blocks corresponding to 4 × 4 pixels, and processing for replacing the coefficient data Z (i) in the block is performed. In the following, the processing of the first layer data will be described, but the same applies to the processing of the second layer data.

In the replacement process, the non-zero coefficient C (i) corresponding to the non-zero coefficient forward coordinate i which is the arrangement order in the VLC code of the non-zero coefficient data, which is variable-length decoded by the VLC decoding unit 120, and the non-zero coefficient Of the zero run length R (i) representing the number of zero coefficients preceding the zero coefficient C (i), the non-zero coefficient C (i) is the coefficient position when the immediately preceding non-zero coefficient data includes the zero coefficient. J _PRE + R (i) +1 obtained by adding the zero run length R (i) and 1 to the coefficient position coordinate j _PRE (the initial value for the first non-zero coefficient in the block is 0) as a new coefficient position coordinate j As the coefficient data Z (P (j)), the coefficient positions are switched and output. The replaced Z (P (j)) is represented as Z ′ (j). Further, although not generated as intermediate data, coefficient data including a zero coefficient before replacement is represented as Z (j). In the frequency update process for updating the non-zero frequency data after the coefficient replacement process, the non-zero frequency data F (j) stored in the replacement information storage unit 40 is read, and the read non-zero frequency data F (j ) Is updated by adding 1, and the updated non-zero frequency data is F ′ (j). Then, the comparison / replacement process for replacing the replacement data P (j) and the frequency update data F ′ (j) is the coefficient data of the coefficient position coordinate j to which the non-zero coefficient data C (i) is to be transferred. Is performed using Z (j) and replacement data P (j) for replacing coefficient data Z (j) in the block. The replacement data P (j) is set in advance according to the prediction mode in the DC / LP / HP inverse prediction processing unit 140, and the prediction data derived in the DC / LP / HP inverse prediction processing unit 140 is used. Corresponding replacement data P (j) is used. Further, the replacement data P (j) and the non-zero frequency data F (j) are set in advance with initial values. The initial values of the replacement data P (j) and the non-zero frequency data F (j) are the same as those illustrated in FIG. These replacement data P (j) and non-zero frequency data F (j) are stored in the replacement information storage unit 40.

  In FIG. 2, the DC / LP / HP inverse prediction processing unit 140 inputs information on the prediction mode, and the VLC decoding unit 120 inputs the first layer data to the coefficient data replacement processing unit 20 one block at a time. . The coefficient data replacement processing unit 20 to which the zero coefficient data C (i) and the zero run length data R (i) are input outputs a new coefficient position coordinate j to the replacement information storage unit 40 and the replacement information storage. The value of the replacement data P (j) read from the unit 40 (which is an initial value in the case of the first block) is used as an input address, and the non-zero coefficient data C (i) is used as input data in the coefficient data storage unit 30. Output.

Specifically, in FIG. 4, the coefficient data replacement processing unit 20 to which the non-zero coefficient data C (i) and the zero run length data R (i) are input holds the non-zero coefficient data C (i) as coefficient data. Output as input data to be stored in the coefficient data storage unit 30 via the unit 21, and output zero run length data R (i) to the coefficient position coordinate update unit 23 via the replacement processing control unit 22, The coefficient position coordinate updating unit 23 adds the zero run length R (i) and 1 to the coefficient position coordinate j _PRE which is the coefficient position when the immediately preceding non-zero coefficient data includes a zero coefficient, j _PRE + R ( i) +1 is output to the replacement information storage unit 40 as a new coefficient position coordinate j. Note that the updated j is assigned to j _PRE to prepare for the next non-zero coefficient processing. The replacement information storage unit 40 outputs the replacement data P (j) of the new coefficient position coordinate j to the coefficient data storage unit 30 through the coefficient address holding unit 24 as a data address. The block completion signal to which the block decoding completion signal is input from the VLC decoding unit 120 after outputting the m-th nonzero coefficient data C (m) and the zero run length data R (m), which are the total number of nonzero coefficients of the block The generation unit 25 outputs the block replacement completion signal to the coefficient data storage unit 30 after the single coefficient replacement completion signal is output from the replacement processing control unit 22.

  The coefficient data storage unit 30 stores the non-zero coefficient data C (i) as data in the coefficient data Z (P (j)) with the replacement data P (j) as the address position. The replaced Z (P (j)) is represented as Z '(j), and is not generated as intermediate data, but the coefficient data including the zero coefficient before the replacement is represented as Z (j). The coefficient data storage unit 30 that has received the block replacement completion signal from the coefficient data replacement processing unit 20 outputs the replaced coefficient data Z ′ (j) to the DC / LP / HP inverse prediction processing unit 140 as parallel data. To do. The coefficient data Z (j) and Z ′ (j) before and after the replacement are the same as those illustrated in FIG. 18, and the generated first hierarchy data is the same as that illustrated in FIG.

  Next, the coefficient data replacement processing unit 20 starts a single coefficient decoding start signal for starting decoding of the next non-zero coefficient data simultaneously with the start of replacement of one Z ′ (j). Output to. The VLC decoding unit 120 to which the single coefficient decoding start signal is input starts decoding the next non-zero coefficient. Further, after the replacement of one Z ′ (j) is completed, the coefficient data replacement processing unit 20 outputs a frequency update start signal for starting the frequency update processing to the frequency update processing unit 50 and updates it. The new coefficient position coordinate j is output to the replacement information storage unit 40, and then the replacement information storage unit 40 outputs the non-zero frequency data F (j) corresponding to the new coefficient position coordinate j to the frequency update processing unit. Output to 50.

  The frequency update processing unit 50 adds 1 to the non-zero frequency data F (j) input from the replacement information storage unit 40, and the comparison / replacement processing unit 60 and the replacement as non-zero frequency data F ′ (j). The information is output to the information storage unit 40. The replacement information storage unit 40 stores non-zero frequency data F ′ (j) and updates it as F (j) for processing of the next block.

  The table data after the addition is the same as that illustrated in FIG. The frequency update processing unit 50 outputs the updated non-zero frequency data F ′ (j) subjected to the addition process to the comparison / replacement processing unit 60 for each non-zero coefficient.

  The comparison / replacement processing unit 60 compares the non-zero frequency data F ′ (j−1) read from the replacement information storage unit 40 with F ′ (j) input from the frequency update processing unit 50, and F ′ When (j) is larger than F ′ (j−1), F ′ (j−1) and F ′ (j) are replaced, and F ″ (j−1) = F ′ (j), F ′ (j) = F ′ (j−1), and replacement data P (j−1) and P (j) before and after read from the replacement information storage unit 40 are replaced by P ′ A process of (j−1) = P (j) and P ′ (j) = P (j−1) is performed. The update data after the update process is the same as that illustrated in FIG.

  The table data F (j) and P (j) stored in the replacement information storage unit 40 are respectively output F ″ (j−1) and F ″ (j) from the comparison / replacement processing unit 60. And P ′ (j−1) and P ′ (j).

  The adaptive scanning unit 10 performs the above-described processing, and generates first-layer data in which coefficient data is replaced according to non-zero frequencies and preset replacement data. Similar processing is performed on the second layer data, and second layer data in which the coefficient data is replaced is generated. The third layer data is output to the DC / LP / HP inverse prediction processing unit 140 as it is.

  FIG. 5 is a flowchart showing the operation of the adaptive scanning unit 10, and FIG. 6 is a flowchart showing the operation of the coefficient replacement process in the adaptive scanning unit 10. At the beginning of the block processing, after initializing by assigning 0 to Z (1) to Z (15) in S1, non-zero coefficient data in C (i) is obtained by variable length decoding in the VLC decoding processing in S3. , R (i) obtains zero run length data. In the first embodiment, since the process in the adaptive scan unit is shorter in time than the process in the VLC decoding unit (can be processed with a smaller number of cycles), the step of waiting for the end of the coefficient replacement process in S5 Does not exist. Next, at the same time as instructing the start of the coefficient replacement process at S6, the process returns to the VLC decoding process at S3, and the processes from S3 to S6 are repeated for the non-zero coefficients in the block. The coefficient data replacement process performed in parallel with the VLC decoding process uses the non-zero coefficient data C (i) and the zero-run length data R (i) acquired in S3, and coefficient data Z (j) in S21. The non-zero frequency data is updated in S22, and F ′ (j) and F ′ (j) are updated based on the magnitude comparison of the non-zero data F ′ (j) and F ′ (j−1) updated in S24 to S26. A replacement process with P (j) is performed.

FIG. 7 shows the operation of the adaptive scanning unit 10 in a state transition diagram. After the processing starts, the state 0 of the standby ST0 transitions to ST1a for processing of the next block, and ST1a In state 1a, a non-zero coefficient for one coefficient and a zero run length are obtained by VLC decoding. In the first embodiment, since the processing in the adaptive scanning unit is shorter in time than the processing in the VLC decoding unit (can be processed with a smaller number of cycles), in the state 1a in the first embodiment, , Without waiting for completion of state 1b, after the VLC decoding process is completed, if it is not the final non-zero coefficient of the block, transition to state 1a and state 1b, and if it is the final non-zero coefficient of the block, Transition to 1b and state 0. In the state 1b, after the coefficient replacement process, the frequency update process, and the comparison / replacement process are performed for each coefficient, the process ends.

  FIG. 8 is a timing chart showing the operation of the adaptive scanning unit 10. This timing chart is an example when the number of non-zero coefficients is three, and the processing time of each process (the number of cycles required for the process) varies depending on the circuit configuration. In FIG. 8, since the coefficient replacement process i1, the frequency update process k1, and the comparison and replacement process h1 are processed in parallel with the next VLC decoding process vlc2 after the first VLC decoding process vlc1, the processing time efficiency is improved. The number of processing cycles required for decoding can be reduced. In the example of FIG. 8, the processing time of one block is 23 cycles, and it can be seen that the number of processing cycles is smaller than the number of cycles of 33 in the conventional configuration example of FIG.

{Second Embodiment}
As shown in FIG. 3, the adaptive scanning unit 10 according to the second embodiment includes a coefficient data replacement processing unit 20, a coefficient data storage unit 30, a replacement information storage unit 40, a frequency update processing unit 50, and a comparison. As shown in FIG. 4, the coefficient data replacement processing unit 20 includes a replacement processing unit 60. The coefficient data replacement processing unit 20 includes a coefficient data holding unit 21, a replacement processing control unit 22, a coefficient position coordinate update unit 23, a coefficient address holding unit 24, and a block. The completion signal generation unit 25 is configured.

  The adaptive scan unit 10 according to the second embodiment is substantially the same in configuration as the adaptive scan unit 10 according to the twelfth embodiment shown in FIG. 2, but in FIG. A single coefficient replacement completion signal indicating that the comparison / replacement processing for one coefficient in the comparison / replacement processing unit 60 is completed is output from the replacement processing unit 60 to the coefficient data replacement processing unit 20, and in FIG. Only the single coefficient replacement completion signal indicating that the comparison / replacement processing for one coefficient in the comparison / replacement processing unit 60 is completed is output from the replacement processing unit 60 to the replacement process control unit 22 in the twelfth implementation. This is different from the adaptive scanning unit 10 according to the embodiment.

  The configuration of the adaptive scan unit 10 according to the second embodiment has variations in the number of cycles required for the decoding process when the number of cycles required for the decoding process is smaller than the number of cycles of the process in the adaptive scan unit. 3 is a configuration example when there are cases where the number of processing cycles in the adaptive scanning unit may be lower, and after completion of the one-coefficient decoding process in the VLC decoding unit 120, the coefficient data replacement processing unit 20 in FIG. 4, the replacement processing control unit waits for a single coefficient replacement completion signal indicating the completion of the comparison / replacement process in the comparison / replacement processing unit 60, and after the output of the single coefficient replacement completion signal, starts the single coefficient decoding to the VLC decoding unit Output a signal. Thereby, even if the processing time in the decoding process is shorter than the processing time in the adaptive scanning unit, the adaptive scanning unit 10 waits for the end of the process in the comparison / replacement processing unit, The start of the process can be instructed, and the decryption process can be performed normally.

  The operation of the adaptive scan unit 10 according to the second embodiment will be described based on the flowchart of the operation of the adaptive scan unit 10 in FIG. The basic operation of the adaptive scan unit 10 according to the second embodiment is the same as that according to the first embodiment, but in the first embodiment, S4 and S5 in FIG. Although the end of the coefficient replacement process is waited, the first embodiment is different in that there is no step of waiting for the end of the coefficient replacement process of S4 and S5 in FIG. In the adaptive scan unit 10 according to the second embodiment, even when the processing time in the decoding process is shorter than the processing time in the adaptive scan unit by waiting for the end of the coefficient replacement process, The decryption process can be performed normally.

  The operation of the adaptive scan unit 10 according to the second embodiment will be described based on the state transition diagram of the operation in the adaptive scan unit 10 of FIG. The basic state transition of the adaptive scanning unit 10 according to the second embodiment is the same as that according to the first embodiment, but in the first embodiment, the state 1a in FIG. After the VLC decoding process is completed, the state transitions to the state 1a and the next VLC decoding process is started, whereas in the second embodiment of FIG. 7, it is different to wait for the completion of the replacement process in the state 1a of FIG. . In the adaptive scanning unit 10 according to the second embodiment, even when the processing time in the decoding process is shorter than the processing time in the adaptive scanning unit, by waiting for the end of the coefficient replacement process, The decryption process can be performed normally.

  According to the image decoding apparatus 1 of the first and second embodiments having the above-described configuration, first, image encoded data for one image is input from the outside to the bit stream analysis unit 110, and this image The encoded data is analyzed and separated, and the first layer data (HP coefficient), second layer data (LP coefficient), and third layer data (DC coefficient) after separation are input to the VLC decoding unit 120.

The VLC decoding unit 120 performs VLC (Variable Length Coding) decoding on the first layer data, the second layer data, and the third layer data, which are the image encoded data separated by the bit stream analysis unit 110. A variable-length decoding process, which is a process, is performed, non-zero coefficient data and zero-run length data are extracted and output to the adaptive scan unit 10.

  In the adaptive scan unit 10, the first hierarchical data and the second hierarchical data in which the pixel order is replaced by a method determined at the time of compression, the coefficient data is replaced for each block by a procedure reverse to the compression, The first layer data and the second layer data after processing are output to the DC / LP / HP inverse prediction processing unit 140. The adaptive scanning unit 10 executes the VLC decoding process in parallel with the coefficient data replacement process, the frequency update process, and the comparison / replacement process. Therefore, according to the adaptive scanning unit 10 of the first and second embodiments, it is necessary to stop the VLC decoding process during the coefficient data replacement process, the frequency update process, and the comparison / replacement process. Compared to the conventional adaptive scanning unit 130, the processing time can be included in the VLC decoding process period, and the coefficient data replacement process, frequency update process, and comparison / replacement process can be performed efficiently. it can. Note that the third layer data input from the VLC decoding unit 120 is output to the DC / LP / HP inverse prediction processing unit 140 as it is.

  In the DC / LP / HP inverse prediction processing unit 140, the first layer data, the second layer data, and the third layer data for each block input from the adaptive scan unit 10 are respectively between adjacent blocks. The prediction mode is derived by comparing the coefficients of, and the value of the corresponding pixel between the highly correlated blocks is added to the target pixel according to the set arithmetic expression for the prediction mode. Is restored (inverse prediction processing). Then, the processed first layer data, second layer data, and third layer data are output to the inverse quantization unit 150 for each block.

  The inverse quantization unit 150 multiplies the first layer data, the second layer data, and the third layer data, which have been subjected to the inverse prediction processing by the DC / LP / HP inverse prediction processing unit 140, with respective set quantization parameters. Then, the inverse quantization process is performed, and the processed first layer data, second layer data, and third layer data are output to the frequency inverse transform unit 160 for each block.

  The frequency inverse transform unit 160 sequentially processes the image data inversely quantized by the dequantization unit 150 in a raster direction with 4 × 4 pixels as one block, and generates coefficient data for each frequency band as pixel data. Process to convert to. The frequency inverse transform unit 160 converts the first layer data, the second layer data, and the third layer data separated for each frequency band into color components (Y / U / V or Y / U / V / K). Each time, the data is returned to one block (4 × 4 pixels) of two-dimensional space data and output to the color space inverse transform unit 170.

  The color space inverse transform unit 170 processes the image data for each color component input from the frequency inverse transform unit 160 for each pixel, and when the color space inverse transform is not necessary, the color space inverse transform is not performed. When reverse conversion is necessary, if the color component is Y / U / V, it is converted to RGB data, and if the color component is Y / U / V / K, it is converted to CMYK data, and the processed pixel The data is output to the outside as image data for one image.

  As described above in detail, according to the image decoding apparatus 1 of the present example, it is necessary to stop the VLC decoding process during the coefficient data replacement process, the frequency update process, and the comparison / replacement process. Compared with the adaptive scanning unit 130, the processing time can be included in the VLC decoding process, so that the coefficient data replacement process and the replacement information data update process can be performed efficiently. As a result, the processing speed of the entire image decoding apparatus 1 can be increased compared to the conventional case, and an efficient decoding process can be realized. In addition, even when the processing time in the decoding process is shorter than the processing time in the adaptive scan unit, a signal for starting the decoding process can be output after waiting for the comparison / replacement processing unit to end. The decoding process can be normally performed.

  Although the first and second embodiments of the present invention have been described above, the specific modes that the present invention can take are not limited to these. For example, in the above example, the coefficient data replacement processing unit outputs a single coefficient decoding start signal instructing the VLC decoding unit to start processing of one coefficient at the same time as the start of the coefficient data replacement process in the coefficient data processing unit. However, the present invention is not limited to this, and it may be output when the coefficient data replacement process in the coefficient data processing unit is completed, and if possible, it is output at the end of the frequency update process in the frequency update processing unit. You may do it.

DESCRIPTION OF SYMBOLS 1 Image decoding apparatus 10 Adaptive scanning part 20 Coefficient data replacement process part 21 Coefficient data holding part 22 Replacement process control part 23 Coefficient position coordinate update part 24 Coefficient address holding part 25 Block replacement completion signal 30 Coefficient data storage part 40 Replacement information Storage unit 50 Frequency update processing unit 60 Comparison / replacement processing unit 100 Image decoding device 110 Bit stream analysis unit 120 VLC decoding unit 130 Adaptive scanning unit 131 Non-zero coefficient storage / coefficient total number accumulation unit 132 Coefficient data replacement processing unit 133 Coefficient data storage unit 134 Replacement information storage unit 135 Frequency update processing unit 136 Comparison / replacement processing unit 140 DC / LP / HP inverse prediction processing unit 141 Non-zero coefficient total accumulation unit 142 Coefficient output control unit 143 Non-zero coefficient storage / output unit 144 Zero run length storage / output unit 145 Coefficient data Holding unit 146 Replacement processing control unit 147 Coefficient position coordinate update unit 148 Coefficient address holding unit 149 Block completion signal generation unit 150 Inverse quantization unit 160 Frequency inverse transformation unit 170 Color space inverse transformation unit ST0 State 0
ST1 State 1
ST1a State 1a
ST1b State 1b
ST2 State 2

Claims (5)

A decoding unit that decodes the compressed image encoded data to generate non-zero coefficient data C (i) and zero-run length data R (i) of each non-zero coefficient forward coordinate i in block units;
The coefficient position coordinate j of the block generated from the zero run length data R (i) and the coefficient data of the coefficient position coordinate j to which the non-zero coefficient data C (i) is to be transferred, Z (j) is non- Non-zero frequency data F (j) representing the relationship with the frequency that was zero, and replacement data P (j) for replacing the coefficient data Z (j) in the block, the current coefficient position coordinate j And the block decoded by the decoding unit using the replacement data P (j) representing the relationship between the coefficient position coordinate j ′ of the coefficient position coordinate j and the coefficient position coordinate j ′ to be incorporated by replacement. An adaptive scan unit that replaces the unit non-zero coefficient data C (i) in the block;
An inverse quantization unit that inversely quantizes the n coefficient data in units of blocks replaced by the adaptive scan unit;
In an image decoding apparatus comprising at least a frequency inverse transform unit that inversely frequency transforms the inversely quantized coefficient data in units of a predetermined number of pixels to generate image data,
The adaptive scan unit includes:
A replacement information storage unit for storing the non-zero frequency data F (j) and replacement data P (j);
Coefficient data Z (j) corresponding to the non-zero coefficient data C (i) and zero run length data R (i) decoded by the decoding unit is replaced with replacement data P (j stored in the replacement information storage unit. ) Based on the coefficient data replacement processing unit for replacement processing,
Based on the replacement processing information of the replacement data P (j) and the non-zero coefficient data C (i), the non-zero coefficient data C (i) is stored in Z (P (j)) in the block. A coefficient data storage unit that performs replacement processing and outputs it as Z ′ (j);
The non-zero frequency data F (j) is updated by adding 1 to the non-zero coefficient position coordinate j of the non-zero coefficient position coordinate j after the decoding, and the updated non-zero frequency data F ′ (j ) Frequency update processing unit,
Using the non-zero frequency data F ′ (j) updated by the frequency update processing unit, F ′ (j−1) and F ′ (j) are sequentially compared, and F ′ ( j −1) and Based on the comparison result of F ′ ( j ), a comparison / replacement processing unit for performing replacement processing of F ′ ( j− 1) and F ′ ( j ), P ( j− 1) and P ( j ), A coefficient data replacement processing unit configured to output a single coefficient decoding start signal for starting decoding of the next non-zero coefficient in parallel with the replacement processing from the coefficient data replacement processing unit to the decoding unit; A frequency update start signal is output to the frequency update processing unit after the coefficient data replacement processing. However, n is an integer of 2 or more, i is an integer of 1 to n or less, and j and P (j) are integers of 1 to n.   2. The image decoding according to claim 1, wherein the adaptive scanning unit outputs a single coefficient replacement completion signal indicating that the comparison / replacement processing is completed from the comparison / replacement processing unit to the coefficient data replacement processing unit. Device.   The image decoding apparatus according to claim 1, wherein the adaptive scanning unit outputs the single coefficient decoding start signal for starting decoding of the next non-zero coefficient simultaneously with the start of the replacement process. . The coefficient data replacement processing unit
A coefficient data holding unit that outputs the non-zero coefficient data C (i) input from the decoding unit to the coefficient data storage unit as input data to the storage unit;
A replacement process control unit for controlling the start of the frequency update process and the start timing of the decoding process;
A coefficient position coordinate update unit for generating a current coefficient position coordinate j;
A coefficient address holding unit that obtains replacement data P (j) from the replacement information storage unit using the current coefficient position coordinate j, and outputs it to the coefficient data storage unit as an input address to the coefficient data storage unit;
A block that outputs a block replacement completion signal indicating completion of block replacement to the coefficient data storage unit based on a block decoding completion signal input from the decoding unit and a single coefficient replacement completion signal input from the replacement processing control unit A completion signal generator,
The replacement processing control unit outputs a single coefficient decoding start signal for starting decoding of the next non-zero coefficient to the decoding unit, and outputs a frequency update start signal to the frequency update processing unit after the coefficient data replacement processing. The image decoding apparatus according to claim 1, wherein the image decoding apparatus outputs the image decoding apparatus. The coefficient data replacement processing unit
5. The image decoding apparatus according to claim 4, wherein a single coefficient replacement completion signal indicating completion of replacement processing is input from the comparison / replacement processing unit.

Images