JP5260425B2 - Image decompression apparatus, method and program - Google Patents

Description

  The present invention relates to a technique for decoding compressed image data in which image data is divided into a plurality of regions and individually compressed.

  In Patent Document 1, variable length decoding processing is performed on image data compressed by JPEG, and orthogonal transform coefficients are obtained (step 31). Also, the resolution (size) of the compressed image data is acquired (step 32), and some orthogonal transform coefficients to be used in inverse transform are determined from the orthogonal transform coefficients based on the resolution of the target reduced image and the acquired resolution. (Step 33). The determined orthogonal transform coefficient is inversely transformed to obtain a first reduced image (step 34). If the resolution of the obtained first reduced image is not the target, it is assumed that the data amount of the reduced image data is adjusted and the target second reduced image is obtained (step 35).

  In Patent Document 2, compressed data having a data structure in which a restart interval marker is inserted for each block unit to be compressed is created. As a result, the DC value at the head of the compressed data of each block becomes an absolute value, and each restart interval marker specifies the recording position in the flash ROM only with address information (without bit information). Therefore, it is not necessary to specify the bit position when reading an arbitrary block from the flash ROM, and since the DC value at the beginning of another block is not required when decompressing with the JPEG decompression block, the block alone does not require that block. Suppose that the compressed data can be decoded.

  In Patent Document 3, it is assumed that image data corresponding to the minimum resolution necessary for image processing on the server computer side is transmitted from the client computer to the server computer.

  The image processing apparatus of Patent Document 4 is selected with a dividing unit that divides the acquired image information into a plurality of blocks, a significant block selecting unit that selects a significant block including characters from the plurality of divided blocks. A compression unit that compresses a significant block at one of a plurality of compression ratios and outputs compressed data; a decompression unit that decompresses compressed data and outputs a decompressed image; and when the entire image information is compressed at a plurality of compression ratios An expected file size calculation unit for calculating each expected file size.

  FIG. 9 is a flowchart showing an outline of a conventional JPEG decoding process. The JPEG decoding process is performed by repeating Huffman decoding (S21), inverse quantization (S22), and inverse DCT (S23) for each decompressed block (S24). A variable length code of one block includes a DCT coefficient including one DC coefficient and 63 AC coefficients. Note that JPEG2000 decoding processing is performed in the same flow as in FIG. 9, and block-by-block EBCOT (Embedded Block Coding with Optimal Truncation) decoding, inverse quantization, and inverse discrete wavelet transform correspond to S21 to S23, respectively.

  The sign of the DC coefficient is determined by a code table as shown in FIG. 10, for example. The code table itself is determined at the time of JPEG encoding and is described in the header of the JPEG file.

  The DC coefficient takes an integer value from −255 to 255. This value is classified into 9 groups. For example, if the DC value is 50, the classification is “6” according to the table. At this time, the binary number “1110” representing the classification “6” is encoded, and is 50 out of the values of −63 to −32 and 32 to 63 included in the classification “6” by 6 additional bits. Specify that. Since the classification “6” includes the possibility of 64 kinds of DC values, it is possible to specify the DC value by adding 6 bits after the “1110” in 2 ^ 6 = 64.

  The AC coefficient is determined by a code table as shown in FIG. 11, for example. This table is also determined at the time of JPEG encoding and is described in the header of the JPEG file.

  Similarly to the DC coefficient, the AC coefficient is classified by value. The classification is represented by a value from 0 to 10 and is equal to the number of additional bits. When an AC coefficient is encoded, it differs from the DC coefficient in that it simultaneously represents the value of the coefficient itself and the number (run length) followed by the zero coefficient preceding the coefficient. That is, for example, the three AC coefficients following “0, 0, 4” are represented by a Huffman code corresponding to “classification = 3, run length = 2” and 3 additional bits.

  If the run length is 16 or more at this time, a code (ZRL) of run length = 16 is once output, and then a code is issued based on the number obtained by subtracting 16 from the original run length.

  At the time of decoding, the run length and the number of additional bits = classification are obtained from the above table, and the AC coefficient is decoded using the subsequent additional bits. This procedure is continued until 63 AC coefficients are decoded.

JP 2006-60540 A JP 2006-313984 A JP 2004-128948 A JP 2008-85914 A

  In recent years, with the spread of terminal devices having a small display screen such as a mobile phone, there is a need to output an image on the display screen by reducing the resolution to an arbitrary resolution while cutting out an arbitrary area from the compressed recorded image. .

  For example, consider a case where a recorded image having a large area such as a newspaper is displayed on a display screen of a small terminal device such as a mobile phone. When the entire newspaper is reduced and displayed on the entire screen, the article is too small to be read, so the part you want to read is cut out from the newspaper and then expanded so that it is suitable for the display screen of the terminal device so that you can read the extracted part It is more realistic to display this at a higher resolution.

  In this case, as in Japanese Patent Application Laid-Open No. H11-260260, if the procedure of once decoding the entire compressed recording image and cutting out the image of the portion necessary for display is performed, decoding of the portion not displayed is useless.

  In Patent Document 2, unless a process of inserting a restart interval marker is performed for each block unit to be compressed, the compressed data cannot be decoded in a single block.

  According to the present invention, in accordance with a normal compression processing technique in which information indicating a block unit is not inserted, the entire compressed recording image is expanded from the compressed image data obtained by dividing the image data into a plurality of regions and individually performing the compression processing. Thus, there is provided a technique for cutting out only an image necessary for data output in various forms such as display, printout, and network transmission without going through the process.

  The present invention includes a range designating unit that accepts designation of a range to be output in a compressed image that is variable-length-encoded in units of blocks, a decoding unit that performs variable-length decoding on the compressed image, and a range accepted by the range designating unit. There is provided an image decompression apparatus comprising: an extraction unit that extracts a block to be decompressed based on a decompression unit that decompresses only a block extracted by the extraction unit from a compressed image that has been subjected to variable length decoding by a decoding unit. .

  A target resolution acquisition unit that acquires the target resolution, and a determination unit that determines a component of the compressed image to be expanded from the block extracted by the extraction unit based on the target resolution acquired by the target resolution acquisition unit. The components of the compressed image determined by the determination unit are expanded.

  The image data is expressed by coefficients obtained by discrete frequency domain transform or discrete wavelet transform in units of blocks, and the decompression unit decompresses from the block extracted by the extraction unit based on the target resolution acquired by the target resolution acquisition unit. The power frequency component is determined, and decompressed by performing inverse quantization and inverse discrete frequency domain transform or inverse discrete wavelet transform on the variable-length-decoded compressed image data sequence corresponding to the determined frequency component coefficient.

  The compressed image data represented by the coefficients is variable-length encoded by Huffman code, EBCOT (Embedded Block Coding with Optimal Truncation) method or arithmetic code.

  The variable length encoded image data includes JPEG (Joint Photographic Experts Group) or JPEG2000 image data.

  An output unit that outputs the image data expanded by the expansion unit is provided.

  The output unit can output the image data expanded by the expansion unit to at least one of a display device, a printer, and a network.

  According to the present invention, an image storage server that transmits an image to an image distribution server in response to a request from an image distribution server that distributes an image requested to be transmitted from a portable terminal outputs a compressed image that is variable-length encoded in units of blocks. A step of accepting designation of a range to be performed; a step of variable-length decoding the compressed image; a step of extracting a block to be decompressed based on the accepted range; and a block extracted from the compressed images subjected to variable-length decoding An image decompression method is provided that performs only the decompression step and the step of transmitting the decompressed image data to an image distribution server.

  The present invention requests an image storage server for an image requested to be transmitted from a mobile terminal, and the image distribution server that receives the image from the image storage server has a range to be output in a compressed image that is variable-length encoded in units of blocks. A step of accepting a designation; a step of variable-length decoding the compressed image; a step of extracting a block to be decompressed based on the accepted range; and decompressing only the extracted block of the compressed image subjected to variable-length decoding An image decompression method is provided that performs the steps of performing and transmitting the decompressed image data to a mobile terminal.

  The present invention relates to a step in which an information processing device accepts designation of a range to be output in a compressed image that has been variable-length encoded in units of blocks, a step of variable-length decoding the compressed image, and decompression based on the accepted range There is provided an image decompression method for performing a step of extracting a block to be performed and a step of decompressing only an extracted block of compressed images subjected to variable length decoding.

  An image decompression program for causing the information processing apparatus to execute the image decompression method is also included in the present invention.

  According to the present invention, only the image blocks necessary for outputting the specified image range (display, network transmission, printing, etc.) are expanded, so that the image expansion process is simplified and the necessary range can be output immediately. .

Schematic configuration diagram of image providing system The figure which illustrated the composition of the portable terminal Diagram showing the structure of a JPEG image file Outline flowchart of decompression processing The figure which shows an example of compression recording image data Detailed flowchart of part of inverse transform processing (S7) of DCT coefficient The figure which shows the other structural example (decompression processing with an image storage server) of an image provision system The figure which shows the other structural example (decompression processing with an image delivery server) of an image provision system Flow chart showing outline of JPEG decoding process The figure which shows an example of the code table of DC coefficient The figure which shows an example of the code table of AC coefficient

<First Embodiment>
FIG. 1 shows a schematic configuration of an image providing system according to a first preferred embodiment of the present invention. This system includes an image storage server 1, an image distribution server 2, a mobile phone 3, and a printer 4. The image storage server 1 and the image distribution server 2 are connected by various networks such as a LAN and the Internet. The image distribution server 2 and the mobile phone 3 are connected by various networks such as a mobile telephone communication network. The mobile phone 3 and the printer 4 are connected by various networks such as infrared communication.

  The image storage server 1 and the image distribution server 2 are configured by a computer, and include a CPU, a memory (RAM, ROM), an HDD, a communication unit, a data input / output circuit, a display circuit, an operation device, and the like.

  When the image distribution server 2 requests the image distribution server 2 to distribute the image via the network based on the distribution request for the desired image from the mobile phone 3, the image distribution server 2 transmits the image via the network. A request is made to the storage server 1.

  When the image storage server 1 receives the request from the image distribution server 2, the image storage server 1 acquires a compressed recording image corresponding to the identification information (hyperlink, image file name, etc.) of the image indicated in the request from the image database 1a. The image is transmitted to the image distribution server 2. The compressed recording image is compressed image data obtained by dividing original image data into a plurality of areas and individually compressing them according to a compression method such as JPEG, JPEG2000, or MPEG (moving image). In the case of JPEG compression, the original image data is subjected to discrete cosine transform (DCT), bit allocation is determined, and quantization is performed. In the case of the baseline method, the quantized data is encoded by Huffman encoding to obtain encoded data. In the case of JPEG2000 compression, original image data is subjected to discrete wavelet transform (DWT), and quantization and entropy coding (entropy coding in units of bit planes) by the EBCOT method are performed.

  The image distribution server 2 transmits the image received from the image storage server 1 to the mobile phone 3. The mobile phone 3 can display all or part of the received image on the display screen 211 or transfer it to the printer 4.

  FIG. 2 is a block diagram showing the internal configuration of the mobile phone 3. The mobile phone 3 is configured so that the whole is controlled by the CPU 230. The CPU 230 includes a RAM 231 that is a volatile memory, a ROM 232 that is a nonvolatile memory, a display unit 233 that includes a display screen 211, an operation key 221, a rewritable ROM 234 that is a rewritable nonvolatile memory, and a power supply unit 235. Is deployed.

  The ROM 232 stores a program executed by the CPU 230, and the rewritable ROM 234 stores a program downloaded from the image distribution server 2 by packet communication, and is stored in the ROM 232 or the rewritable ROM 234. When the program is executed by the CPU 230, the operation of each part of the mobile phone 3 is controlled.

  The RAM 231 is used not only as a work area for a program executed by the CPU 230 but also as a storage area for setting values of the number of pixels of image data handled by the mobile phone 3 and image data.

  The display unit 233 has a display screen 211 and displays an image of the RAM 231 on the display screen 211 in accordance with an instruction from the CPU 230.

  The aspect ratio of the display screen 211 is arbitrary, for example, 4: 3, 16: 9. The resolution of the display screen 211 is arbitrary, for example, 320 × 180, 320 × 240, 640 × 480, 1024 × 768, 1280 × 768, 1920 × 1080, and the like. Information indicating the aspect ratio and resolution of the display screen 211 is stored in the ROM 232.

  The operation key 221 has a role of receiving an operation by the user and transmitting the operation to the CPU 230. When the operation key 221 is operated, the CPU 230 controls each unit according to the operation.

  The power supply unit 235 is loaded with a battery (not shown), and the power from the battery is supplied to the CPU 230 and each part of the mobile phone in response to a command from the CPU 230.

  In addition, the cellular phone 3 includes an antenna 213, a transmission / reception unit 241, a signal processing unit 242, and a call unit 243 as components that perform a telephone function. The call unit 243 includes a microphone 243 a provided inside the earpiece 222 and a speaker 243 b provided inside the mouthpiece 212.

  The transmission / reception unit 241 is a circuit element responsible for transmission / reception of radio waves by the antenna 213, and a signal obtained by the transmission / reception unit 241 by reception of radio waves by the antenna 213 is input to the signal processing unit and subjected to signal processing to receive a call unit 243. Are output as sound from the speaker 243b. The voice picked up by the microphone 243 a of the call unit 243 is transmitted as a radio wave from the antenna 213 via the transmission / reception unit 241 for signal processing by the signal processing unit 242.

  The mobile phone 3 also has a packet communication function, and packet signals received via the antenna 213 and the transmission / reception unit 241 are temporarily stored in the RAM 231 after being subjected to appropriate signal processing by the signal processing unit 242, or In the case of a downloaded program, it is stored in the rewritable ROM 234, and the packet data in the RAM 231 is displayed on the display screen 211 of the display unit 233 or stored in the rewritable ROM by the CPU 230 that has received an instruction from the operation key 221. The executed program is executed. Any data may be received by the cellular phone 3 by packet communication, and examples thereof include compressed recording image data, character data, voice data, and a program.

  The packet communication document or the like created with the operation key 221 is temporarily stored in the RAM 231 at the time of creation, sent to the signal processing unit 242 in response to a transmission instruction from the operation key 221, and subjected to signal processing for transmission. 241 and the antenna 213 are transmitted as radio waves.

  Further, the mobile phone 3 is provided with a photographing unit 251, an image processing unit 252, and an infrared communication unit 253.

  The photographing unit 251 includes a photographing lens 214 and a photographing element 251a. A subject captured via the photographing lens is presented by the photographing element 251a to generate an image signal. The image signal obtained by the photographing element 251a is processed by the image processing unit 252 and converted into digital compressed image data compliant with various standards such as JPEG, JPEG2000, and MPEG, and is temporarily stored in the RAM via the CPU 230 and operated. It is displayed on the display screen 211 of the display unit 233 according to the operation of the key 221 or transmitted to an external device such as the printer 4 via the infrared communication unit 253. The infrared communication unit 253 also has an infrared signal reception function. When the transmitted image data is received by the printer 4, the infrared communication unit 253 communicates from the printer 4 with an acknowledge signal indicating that the image data has been received. Receive.

  As a result, in the cellular phone 3, whether or not the transmitted data is correctly received by the printer 4 depends on whether or not the acknowledge signal can be received within a predetermined time after transmitting the image data to the printer 4. Can be recognized.

  Here, the “mobile phone” is merely an example of various portable information terminals such as a notebook personal computer and a PDA. Further, the image transmitted to the printer 4 via the infrared communication unit 253 is not limited to that derived from the photographing unit 251, but may be image data derived from a compressed recording image received from the image distribution server 2 via the transmission / reception unit 241. Good. Which image is transmitted to the printer 4 can be arbitrarily selected via the operation key 221.

  FIG. 3 shows the structure of a JPEG image file which is an example of compressed recording image data received or photographed by the mobile phone 3.

  First, after an SOI marker (a specific data string of 2 bytes) indicating the start of JPEG data, there is a segment representing a compression parameter called DQT or DHT. At the beginning of the segment, a 2-byte segment length is described after the byte marker indicating the segment type, so that if the segment data is unnecessary, it can be skipped. Some segments may not be required. After several segments, there is an SOS marker, and the rest is the image data body.

  After the SOS marker, image data is stored in units of macro blocks called MCUs. In the JPEG system, the MCU is 8 × 8 pixel units or 16 × 16 pixel units. Specifically, the number of colors and the sampling ratio are described in the SOF segment, and the pixel unit can be determined by referring to the description. For example, when the number of colors is 1, or when the number of colors is 3 and the sampling ratio is 4: 4: 4, the MCU is 8 × 8 pixel units, and when the number of colors is 3 and the sampling ratio is 4: 2: 0. The MCU is a unit of 16 × 16 pixels.

  In FIG. 3, description is made assuming a case where the number of colors of a commonly used combination is 3 and the sampling ratio is 4: 2: 0. At this time, if the number of colors is Y, Cb, and Cr, four 8 × 8 pixel blocks of Y and Cb and Cr are sampled (or reduced) from the macroblock of the original image of 16 × 16 pixels, respectively. The MCU is composed of a total of six block data of two 8 × 8 pixel blocks.

  In the data of each block, there are quantized representative values of 64 DCT (discrete cosine transform) coefficients from 0 to 63 in a zigzag order from a low frequency component (upper left) to a high frequency component (lower right). It has been described. That is, each of the 64 DCT coefficients is divided by a quantization width or a quantization step corresponding to a predetermined quantization table, and each DCT coefficient is quantized by rounding the decimal places of the value.

  Since the quantized DCT coefficients are variable-length encoded (entropy encoded) using Huffman code or the like, blocks that are not necessary for output cannot be skipped in the form of “the block ends after”. . Further, when all the DCT coefficients after a certain number are 0 in the block DCT coefficients, it is not necessary to describe the DCT coefficients after that number by describing the EOB marker.

  EOB is a code meaning that the quantized representative values of the subsequent DCT coefficients in the block are all zero. In the DCT coefficient table obtained by performing DCT, a low spatial frequency DCT coefficient in which deterioration of image quality is likely to be manifested is arranged in the upper left, and a high spatial frequency DCT coefficient in which degradation of image quality is less likely to be apparent toward the lower right. Be placed. Since the quantized representative values are entropy-coded in the zigzag order from the upper left to the lower right, this order is called the zigzag order. Therefore, when entropy decoding is performed in the order of the zigzag order, the quantized representative value representing the DCT coefficient having a lower spatial frequency is restored earlier. Then, EOB is added to the head of the column of zero quantized representative values.

  FIG. 4 shows a schematic flowchart of decompression processing executed by the CPU of the image storage server 1. A program for causing the CPU of the image storage server 1 to execute this processing is stored in a storage medium such as a ROM. This program can be executed by a CPU of a personal computer such as the image distribution server 2.

  In S <b> 1, a request for transmitting a desired compressed recording image is received from the mobile phone 3 via the image distribution server 2. Then, the compressed image file corresponding to the image identification information (hyperlink, file name, etc.) included in the request is acquired from the DB 1a to the RAM.

  In S2, information indicating the output range (a part of a newspaper article, etc.) of compressed recorded image data stored in the RAM is input via the operation device of the image storage server 1 or the operation key 221 of the mobile phone 3. And accept. For example, by transmitting the operation information of the cross key and the confirmation key included in the operation key 221 from the mobile phone 3 to the image storage server 1, the start coordinates and end coordinates of the area to be output from the entire image data area, for example, the upper left corner of the rectangle It accepts specification of the coordinates of and the coordinates of the lower right. In addition, selection of the final output destination of the image (output to the display screen 211, output to the printer 4, or various electronic devices such as other mobile phones connected to the mobile phone 3) may be accepted.

  In S3, a block that needs to be decompressed is extracted from the compressed recorded image data based on information indicating the output range of the image data received from the operation key 221 or the like.

  For example, assume that compressed recording image data as shown in FIG. 5 is stored in the RAM of the image storage server 1. This image is divided into 15 × 10 blocks, and the image data of each block is stored in each MCU. Further, it is assumed that the output range Z is designated. In this case, a covering macroblock group Zcov that is a macroblock group including the output range Z with the minimum area is extracted as a block that needs to be expanded.

  Generally, when the start coordinates are Xs and the end coordinates Ys, and one side of the block is N, the (Xs / N) th in the X direction and the (Ys / N) th in the Y direction (however, rounded down in integer units) The first block is assumed to be the 0th block). Further, when the size of the output range Z, that is, the target resolution is Xw × Yh, the lower right block to be cut is ((Xs + Xw−1) / N) in the X direction and ((Ys + Yh−1) / N in the Y direction. ) Th block Bt.

  Note that the compressed recording image data only needs to be divided in units of blocks, and the application range of this processing is not limited to JPEG. In addition, this process can be applied to both moving images and still images.

  In S4, the Huffman code table and the quantization table used for compression are used to Huffman-decode the DCT coefficients of each block subjected to Huffman coding to obtain one DC coefficient and 63 AC coefficients for each block. For example, based on the code tables in FIGS. 10 and 11, only S21 (Huffman decoding) in FIG. 9 may be executed for all blocks. In other words, the decompression process, that is, S22 (inverse quantization) and S23 (inverse DCT) in FIG. 9 is not performed on all blocks, but only on necessary blocks. Details of the compressed data decompression process for only the necessary blocks will be described later.

  S5 to S8 are the same as in Patent Document 1. That is, in S5, the resolution (number of pixels) of the compressed image data in the RAM is acquired from the header of the image file in which the compressed image data is stored.

  In S6, when the display screen 211 is selected as the output destination, the resolution (number of pixels) of the display screen 211 stored in the ROM 232 of the mobile phone 3 is acquired and set as the target resolution. This target resolution may be included in the image transmission request from the mobile phone 3 to the image storage server 1.

  Based on this target resolution, the frequency component of the DCT coefficient to be subjected to IDCT (Inverse Discrete Cosine Transform) is determined among the blocks included in the cover macroblock group Zcov. As in Patent Document 1, the reduction ratio α = target resolution / original compressed image data resolution is calculated, and among all the DCT coefficients, a number of DCT coefficients corresponding to the ratio of α along the zigzag order from the low frequency side are calculated. Use to generate an image that is α times the original image. For example, when the output range Z is to be output at a target resolution of 3/8 of the original image, the upper left of each 8 × 8 block constituting the MCU included in the cover macroblock group Zcov (in zigzag order from the low frequency side) 3) × 3 coefficients are determined as coefficients to be inversely transformed.

  If a printer is selected as the output destination, the print resolution is acquired from the printer 4 and is set as the target resolution.

  In S7, IDCT is performed only on the DCT coefficients of the frequency components determined in S6 to obtain thinned image data. If the target resolution is smaller than the resolution of the original image, reduced image data is obtained. However, since the display screen 211 of the mobile phone 3 is small, reduced image data will normally be obtained according to the target resolution specified from the mobile phone 3. The inverse transform of the DCT coefficients is partially performed only on the coefficients determined in S6 among the coefficients included in the cover macroblock group Zcov extracted in S3, and all the DCT coefficients of all blocks are inversely transformed. Not to do. This point will be described in detail below.

  In S8, final output data is created by cutting out the portion corresponding to the output range specified in S2 from the thinned image data obtained in S7. If the size of the final output data is not the same as the target resolution, the data amount of the image data may be further adjusted so as to be the target resolution, and this may be changed to the final output data. The adjustment of the data amount can use, for example, bicubic interpolation that calculates and enlarges or reduces the peripheral image of the original image. The final output data is transmitted from the image storage server 1 to the mobile phone 3 via the image distribution server 2. The mobile phone 3 can be used for various purposes such as displaying the received final output data on the display screen 211 and transmitting it to the printer 4.

  FIG. 6 shows a detailed flowchart of S7 of decompression processing (inverse transformation of a part of DCT coefficients). This process includes the DCT coefficients of the blocks corresponding to the covering macroblock group Zcov among the DCT coefficients of each block of each MCU obtained by variable length decoding (Huffman decoding) of the bit stream representing the image data body. IDCT is applied only to some frequency components.

  In S10, 64 units of DCT coefficients (one DC coefficient and 63 AC coefficients) Huffman-decoded from the beginning of the image data main body in zigzag order from the low frequency side to the high frequency side are counted to obtain one unit worth. Identify the block. As described above, in S4, one DC coefficient and 63 AC coefficients are obtained for each block by Huffman decoding. Therefore, if one DC coefficient and 63 AC coefficients are counted, one block is identified. can do. If an EOB marker is detected during the counting, the counting of the DCT coefficient is terminated at that time.

  In S11, whether the covered macroblock group Zcov includes 64 DCT coefficients (DCT coefficients from the lowest frequency DCT coefficient to the EOB when censored by EOB) identified as a block of one unit in S10. Judge whether or not.

  For example, when the number of colors is 3 and the sampling ratio is 4: 2: 0, the MCU is in a unit of 16 × 16 pixels. Therefore, the sixth block identified from the beginning of the image body data is MCU (1), The seventh to twelfth blocks belong to MCU (2)... If the MCU to which a certain block belongs is included in the covering macroblock group Zcov, the block is included in the covering macroblock group Zcov. become. Therefore, it is only necessary to determine whether the block is included in the cover macroblock group Zcov by determining whether the MCU to which the identified block belongs is included in the cover macroblock group Zcov.

  In the case of Yes, it progresses to S12. In the case of No, the process returns to S10, from the next DCT coefficient at the end of the already identified block for one unit (the 64th DCT coefficient located at the highest frequency, or the end of the 0 DCT coefficient following EOB). The blocks for one unit are identified by counting 64 DCT coefficients from the low frequency side to the high frequency side in zigzag order.

  In S12, the IDCT based on the target resolution is applied to the 64 DCT coefficients (or the DCT coefficients from the lowest frequency DCT coefficient to the EOB) determined to be included in the cover macroblock group Zcov based on the target resolution. The frequency component of the DCT coefficient to be subjected to (inverse discrete cosine transform) is determined (S6), and the quantized representative value is subjected to inverse quantization and IDCT only on the determined DCT coefficient of the frequency component to obtain image data (S7). ).

  After the above processing, the process returns to S10, and the next DCT at the end of the already identified block of one unit (the 64th DCT coefficient located at the highest frequency or the end of the 0 DCT coefficient following EOB) From the coefficient, the next one block is repeatedly identified by counting 64 DCT coefficients from the low frequency side to the high frequency side in zigzag order. If all blocks constituting the covering macroblock group Zcov are subjected to inverse quantization and IDCT in S12, the process returns to S8 without returning to S10.

  With the above processing, inverse quantization and IDCT are applied to only a part of the DCT coefficients of the block necessary for achieving the target resolution at the output destination (display, network transmission or printing) for the output range of the image designated by the user. As a result, the image output process is simplified, and the required range can be immediately output at the required resolution.

  In the case of JPEG2000, if JPEG Huffman decoding, inverse quantization, and discrete inverse cosine transform are read as EBCOT decoding, inverse quantization, and inverse discrete wavelet transform, respectively, the output is the same for JPEG2000. Inverse quantization and inverse discrete wavelet transform can be applied only to some discrete wavelet coefficients of a block necessary for achieving the target resolution.

Second Embodiment
The execution subject of the decompression processing in FIGS. 4 and 6 is not necessarily the image storage server 1, and can be changed according to the system configuration and application. 7 and 8 show other configuration examples of the image providing system.

  For example, as shown in FIG. 7, when the image distribution server 2 accepts designation of an arbitrary compressed recording image from the mobile phone 3, the image storage server is configured to acquire and transmit the designated compressed recording image from the image database 1a. 1 and receives the compressed recording image transmitted from the image storage server 1 in response to the request. When the image distribution server 2 receives designation of an arbitrary output range for the compressed recorded image from the mobile phone 3, the image distribution server 2 performs the decompression processing of FIGS. 4 and 6 on the compressed recorded image, and transmits the obtained image data to the mobile phone 3. May be sent to.

  Alternatively, as shown in FIG. 8, when the image distribution server 2 receives designation of an arbitrary compressed recording image from the mobile phone 3, the image storage server is configured to acquire and transmit the designated compressed recording image from the image database 1 a. 1, and the compressed recording image transmitted from the image storage server 1 in response to the request is received and transferred to the mobile phone 3. The cellular phone 3 may perform the decompression processing of FIGS. 5 and 6 on the compressed recording image received from the image distribution server 2 and transmit the obtained image data to the printer 4.

  In any aspect, it is possible to reduce the amount of image data transmitted over the network.

1: Image storage server, 2: Image distribution server, 3: Mobile phone, 4: Printer

Claims (11)

A range designation unit that accepts designation of coordinates of a range to be output in a compressed image that is variable-length encoded in block units;
A selection section for selecting an output destination of the image;
A decoding unit for variable-length decoding the compressed image;
An extraction unit that extracts a block to be expanded based on the coordinates of the range received by the range specification unit and the target resolution corresponding to the output destination of the image selected by the selection unit ;
Of the compressed image decoded by the decoding unit, the decompressing unit that decompresses only the block extracted by the extracting unit;
An image expansion apparatus comprising: A target resolution acquisition unit for acquiring the target resolution;
A determination unit that determines a component of a compressed image to be decompressed from the block extracted by the extraction unit, based on the target resolution acquired by the target resolution acquisition unit;
With
The image decompression apparatus according to claim 1, wherein the decompression unit decompresses a component of the compressed image determined by the determination unit. The compressed image is represented by coefficients obtained by discrete frequency domain transform or discrete wavelet transform in units of blocks,
The expansion unit determines a frequency component to be expanded from the block extracted by the extraction unit based on the target resolution acquired by the target resolution acquisition unit, and performs variable length decoding corresponding to the coefficient of the determined frequency component The image expansion apparatus according to claim 2, wherein the compressed image data sequence is expanded by performing inverse quantization and inverse discrete frequency domain transform or inverse discrete wavelet transform.   4. The image decompression apparatus according to claim 3, wherein the compressed image data expressed by the coefficients is variable length coded by a Huffman code, an EBCOT (Embedded Block Coding with Optimal Truncation) method, or an arithmetic code.   5. The image decompression apparatus according to claim 4, wherein the variable length coded image data includes JPEG (Joint Photographic Experts Group) or JPEG2000 image data.   The image decompression apparatus according to claim 1, further comprising an output unit that outputs image data decompressed by the decompression unit.   The image decompression apparatus according to claim 6, wherein the output unit is capable of outputting the image data decompressed by the decompression unit to at least one of a display device, a printer, and a network. An image storage server that transmits the image to the image distribution server in response to a request from an image distribution server that distributes an image requested to be transmitted from a portable terminal,
Receiving a specification of coordinates of a range to be output in a compressed image that is variable-length encoded in block units;
Selecting the output destination of the image;
Variable length decoding the compressed image;
Extracting a block to be decompressed based on the coordinates of the accepted range and a target resolution corresponding to the output destination of the selected image ;
Decompressing only the extracted block of the variable length decoded compressed image;
Transmitting the decompressed image data to the image delivery server;
An image decompression method that executes An image distribution server that requests an image storage server for an image requested to be transmitted from a mobile terminal and receives the image from the image storage server,
Receiving a specification of coordinates of a range to be output in a compressed image that is variable-length encoded in block units;
Selecting the output destination of the image;
Variable length decoding the compressed image;
Extracting a block to be decompressed based on the coordinates of the accepted range and a target resolution corresponding to the output destination of the selected image ;
Decompressing only the extracted block of the variable length decoded compressed image;
Transmitting the decompressed image data to the mobile terminal;
An image decompression method that executes Information processing device
Receiving a specification of coordinates of a range to be output in a compressed image that is variable-length encoded in block units;
Selecting the output destination of the image;
Variable length decoding the compressed image;
Extracting a block to be decompressed based on the coordinates of the accepted range and a target resolution corresponding to the output destination of the selected image ;
Decompressing only the extracted block of the variable length decoded compressed image;
An image decompression method that executes   An image decompression program for causing an information processing apparatus to execute the image decompression method according to claim 10.

Images