TW395118B - Image encoding method, image encoding device, image decoding method and image decoding device - Google Patents

Abstract

The load of attached information in image signals can be reduced even when the conditions such as bit stream change. Based on the most suitable encoding mode that will not lower the encoding rate, the encoding mode selector 2 of the image-encoding device has two encoding mode table A & B which include the identification number of encoding mode in the encoded image. One of the encoding mode table 1 and 2 is selected according to the encoding mode table. The encoding mode with best encoding rate is selected from several encoding modes in the selected encoding mode table to get the encoded image signal.

Description

No. 86115728 amendment page Λ7 B? ^ Date: 88 · 1 · 26 V. Description of the invention (32) Change the unit% to output the predicted image 219 to the addition unit 7 Add the predicted image 219 to the inverse orthogonal transform unit 7 , On the No. 77 of the addition section 216, and output as the exhaustion image signal mirror 217. ^ Prediction error signal Decoded image region encoded by the internal mode. On the other hand, in the above, the conversion section 76 will rotate out. In the addition section 77, the second and lower decoded prediction error signals of the orthogonal transform section 73 in step ST57 are inverse signal 217. The output is decoded in the step ST58 of the Decoding Video Association. Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs: The code 71 is compiled according to the figure shown in Figure 5. Shadow ::: =: Etc. Every-decoding process 2 == all the encoded data in the case where the encoded data of the image area can be detected and there is also 'Dong Dong Fu from the initial step ST51 due to the decoding of the security guard U AL ^ case ^ * method Detected encoded video encoding data, etc. "Because all the decoded encoded data has not been processed for decoding. This decoded video signal 217 is a f: = ^ device (not _, a later Sr. loss image 218, and is written into the memory 75. Fig. 27 shows the syntax analysis and the operation of the variable-length encoding unit η. That is, the flowchart of the processing of step ST51 of the 26th objective. In the step lamp 61, the π boat selects β decoding in the encoding mode table of FIG. 24. " 1 is to decode the encoding bit stream 213 command selection information 200. Shima. In step ST62, the conversion unit 82 judges whether the coding mode table selection information 200 has been decoded according to the decoded coding mode table selection information 200 '------------ — (Please read the notes on the back before filling this page) 1T ^ ----- _ :: ”: LL_

"IJJ ~~: ______ 35 paper rule M B7 V. Description of the invention (ι) The present invention relates to an image encoding method, an image encoding device, an image decoding method, and an image decoding device, which correspond to the transmission encoding bit & .Stream_ block absolute bit rate and other conditions, from a plurality of pre-scaled multiple encoding modes, Huang_Choose the encoded and decoded video signal or decoded video signal (with * ...________________ ', referred to as coded Video signal) the most suitable encoding mode for encoding ----------- encoding and decoding. _ First, the general video encoding device and video decoding device related to the technical background of the present invention and their problems will be described. FIG. 1 is a diagram showing an example between an image encoding device and an image decoding device.

Schematic diagram of the structure of a coded bit stream that is generally transmitted and received. Q ...-... '… ... Because the coded bit stream is made into a frame (1 day surface) at each time, as As shown in FIG. 1, it is composed of encoded image header information and each piece of encoded data of each region 1 to N when a frame is divided into N regions. The encoded video header information * includes the quantization method selection information when the quantization method is used at the encoding and decoding side. Each coded data of each area 1 ~ N is composed of compressed image data and-乂 > -〆 * · _,-'.-Corresponding to each quantization code. It is composed of object information composed of change information, etc. Therefore, when there are not many types of coding modes, the amount of coding mode information in the coded data decreases, and as a result, the amount of object information also decreases. The most suitable coding mode cannot be obtained through the encoded video signal, so it is not possible to obtain a highly efficient Encoding, resulting in poor day quality. On the other hand, in order to improve the day quality during decoding, the Chinese National Standard (CNS) A4 specification (210X 297 mm) is applied to match the paper size. (Please read the precautions on the back before filling in this page. -HI- · Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs. \ Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7. 5. Description of the invention (2) The encoded video signals are registered in many encoding modes. The amount of information in the stone horse mode of data t increases, and the amount of object information also increases. When the size of the encoded data is determined by a fixed length and a variable length, the amount on the compressed image data side will become smaller, and the day quality will worsen. Therefore, there is a problem of poor coding efficiency ... This problem is particularly obvious when not only the amount of coded data is small but also the bit rate is low. Next, use ISO / IEC JTC1 / SC29 / WG11 Describe the above-mentioned problems in the video coding device and video decoding device, especially when using the current standardization MPEG-4. For video VM8.0 (ISO / IEC JTC1 / SC29 / WG11 / N1796), it is recommended ！ Select one of the 4 microblock coding modes (hereinafter referred to as MBT-YPE). For B-VOP (Bidirectionally Predictive-Video; __________—...

Object Plane). \ ____ · 一 " Figure 2 is a diagram showing the MBTYPE table (MBTVΡΕ-0) for B-VOP coding of VM8.0 proposed here. Here, it is recommended to select one of the four encoding modes 卞 designated by MBTYPpO ~ 4 for encoding. The selected benchmark is not a standardized range, and it is recommended to use prediction error power. In the figure, X indicates data (DQUANT, MVDf, MVDb, MVDB) used in each encoding mode. Details are described in vM8.〇 pp.77, 3_5.5. For this VM8.0 proposal, ‘because the module & complies with 4 releases ~. According to the above reasons, none, encoding of microblocks is performed efficiently, and there is a problem of poor day quality. I ............... As the B-VOP coding method suggested above, if it is too biased to correspond to the low 5 paper standards applicable to Chinese countries ( CNS) A4 specification (210X ^ 97mm) '---------- r'; ^ ------ IT ------ ^ 1®—ί,-. (Please Read the notes on the back before filling this page) A7 B7 V. Description of the invention (3) There are not many 4 coding modes and the flexibility is low. As a countermeasure for this, consider setting more than 10 kinds of encoding modes for each microblock of the B-VOP '— · —-·· " ..... ..- · ... However, at this time, 'Although the softness during encoding is increased, the number of bits used to express each -----.......-, .....' Increase, when low bit rate encoding, the influence on the increase of object information becomes larger, which makes the encoding efficiency worse. In particular .................._ ------....-- First, in the absence of prediction, but only for a temporary representation, that is, with sufficient B_v〇p, the DCT coefficient occupies The ratio should be more than i_v〇p (Intra_vide〇〇bjeet

Plane) or p_v〇P (Rredictive-Video Object Plane) is low. For B-VOP, only the proportion of object information in this part will be higher, and the coding efficiency will be lower. In order to solve the above problems, an object of the present invention is to provide an image encoding method, an image encoding device, an image decoding method, and an image depletion device, that is, a unit rate and the like. When the conditions are changed, also: i: Wei Yanxian, who listens to the silk, chooses the most suitable mode for encoding and shovel. 、， 一 — 、. Optimized and another 'Another object of the present invention is to provide an image image encoding device, an image decoding method, and an image _ device. When conditions such as code set and bit rate of Guang Cao ’s video signal transmission when the video decoding device is changed; can be healthy and decent (burden, will not cause editing effects *, which means that the number of β minus U can be registered first. Select the best encoding mode in the encoding mode for encoding and decoding .. Code, front (Please read the precautions on the back before filling this page} The video encoding method of the present invention is in the specified area Previous-* ·· * —— — This paper size is applicable to the Chinese national standard CNS) ^ 4 娜 (210 ^ 7 ^ Five printed by the Central Bureau of Standards of the Ministry of Economy Staff Consumer Cooperatives A7 B7 Description of invention (4) One side of coding mode For the encoding of the input video, you can select the encoding according to the encoding mode group selection information in the complex encoding mode group that can select the encoding mode for each area from the registration ~ ·--1. Pattern group The selected coding mode group encodes the input image in the prepared area to generate a coded bit stream, and adds the above coding mode group selection information to the coded bit stream and outputs it to the code bit stream. ·· ~.... According to this, even if the encoding mode increases, the amount of encoding mode information will not-_ _ · -... increase, so Kosima can perform efficient encoding of image encoding Method: The image encoding device related to the present invention includes a plurality of encoding modes in a predetermined area, which are converted and edited on the input side, and are inputted into an input image encoding mode. The coding mode group selects the information and selects the coding mode group. The coding mode group of the coding mode group — " '* · — ...... — ~ _ Select a device to log in to the selected program. The coding mode of the exhibition group Φ, choose to assemble the house basket in the above areas....... The selected encoding mode outputs an encoding device that encodes the encoded data of each area, so that the encoding mode Group selection information, above • —........-............ ............. Coding The mode and the above-mentioned encoding data are multiplexed and output as an encoded bit stream r_ '' '................. … Multiplexing device. According to this, even if the encoding mode is increased, the amount of encoding mode information will not increase, so an image encoding device that can perform efficient encoding can be realized. — The image related to the present invention Encoding device, fish boil "solid" peak obstructive model Wu Wuxian first _ login to select the encoding mode of the punishment block as the unit of the encoded image, the paper size of the edition applies the Chinese National Standard (CNS) A4 specification (210X297 mm) (Please read the notes on the back before filling in this I-) Order-line printed by the Consumer Standards Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7. 5. Description of the invention (5) Code. The group selection device is used to display the video object at each moment. The state of the video and the video object plane composed of the above micro and block, according to the coding mode group selection information, select the coding mode for coding the _ code from the plurality of coding mode groups. Group. 'So' Even if the encoding mode I is strong, the amount m of the u-bin type is increased: so that it can realize MPEG-4 pairs that can perform efficient encoding —--------------- ----- -------- „-· ..--·-. Appropriate image encoding device. Regarding the gray image encoding device of the present invention, the $ code device includes a plurality of different quantums. Device, select the information according to the quantization method, and select the above ... '· —___—.. ..........-............. ..—........-_. _..- 〆 · Quantization Selector of Any of the Quantizers' When encoding, use the selected quantizer to perform quantization, And the coding group selection device uses the above quantization method selection information as the coding mode group selection information. In this way, even if the coding mode is increased, the amount of coding mode information will not increase, and it can be implemented efficiently under the given conditions. At the same time, due to the existing quantization method of Xiangdan, the information is selected as the coding mode group. _________---- Information, 5 * to implement a highly efficient encoding $ image encoding device. '' ' ..-----... — " * *-Regarding the image coding device of the present invention, a plurality of coding mode groups are defined by r ', ........... The encoding mode corresponding to the bit rate corresponding to the low bit rate is registered as the encoding mode group corresponding to the bit rate corresponding to the encoding mode group, and the encoding mode corresponding to the bit rate higher than the above reference bit rate The registered high bit rate corresponds to the coding mode group. In this way, it is possible to realize an image coding device that can correspond to a wide range of bit rates from low bit rates to high bit rates. 8 This paper standard applies Chinese National Standards (CNS) A4 specification (210X297 mm) (Please read the notes on the back before filling in this page) Λ Order A7 A7 Printed by the Central Standards Bureau of the Ministry of Economic Affairs, Printed and Invented by the Cooperative Consumer Cooperative (6) The image decoding method related to the present invention is In the method of editing the bit stream of the stone and encoding the image into a predetermined region, the method of decoding the image is decoded, and the region I code bit stream decodes the encoding mode group selection information. The coding mode of the cooked farm news ^ should be the selected coding mode: Decode the encoding data of each mater in the * coded bit A stream. Second, it is expected that the amount of coding mode information and the status of the coded bit stream can be decoded positively under the given conditions A decoding method that can perform efficient decoding can be realized. 2 The video of the present invention includes the compression of a bit stream in a private image, and in a predetermined area and image, it can be edited at the time. The plurality of codes of the fine coding mode * 丄 士 群 'selects a coding mode group selection information decoding device that decodes the coding mode group selection information from the above coding, and selects the coding mode group selection information from the plurality of coding mode groups by using the coding mode. The touch mode group selection device that selects the indicated coding mode group from the information, and uses the coding bit property of the selected mode group to use the coding ^ in the above-mentioned t region! Editing means — According to the encoding mode, a decoding means that decodes the encoded data of each region from the encoded bit stream. In this way, even in the case of a coded bit stream with a small amount of coding mode information, it can be correctly decoded. Under the given conditions, a decoding device capable of efficient decoding can be realized. This paper is based on the specifications of the former middle and first quasi-A4 (210X297): (Please read the precautions on the back before filling this page) --- 1 S-IT ------------ ^ TIIL --- Printed by A7 B7 of the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the Invention (7) Regarding the image decoding device of the present invention, a plurality of coding modes are related to 4 series, .....--- ----_ 'Register the coding mode of the microblock that may be selected as the unit of the coded image first. ~~ 一 * ---------- One-one coding mode group selection device is indicated by On the video object plane composed of a plurality of the above-mentioned microblocks of the image of the state of the video object at each moment, according to the coding mode group selection information, select from a plurality of coding mode groups for use in " --- --.... ........... Coding mode group for encoding. \ _ 1 ........ In this way, even in the case of a coded bit stream with a small amount of coding mode information, it can be correctly decoded. Under these conditions, an encoding device corresponding to MPEG-4 that can perform efficient encoding can be realized. With regard to the image decoding device of the present invention, the decoding device includes a plurality of ... &-quot;.--............- — —— · ———— ~ different inverse quantization devices, Select the information according to the quantization method, and select -〆- · _ — .................... .... * —— _____________——— An inverse quantization selection device for any of the inverse quantization devices. In decoding, the selected inverse quantization device is used for inverse quantization, and the encoding mode group selection device uses the above quantization method selection information as the encoding mode group selection information. In this way, even in the case of a coded bit stream with a small amount of coding mode information, it can be correctly decoded. Under the # given by Wei, efficient decoding can be implemented. On the same day, the existing quantization method was selected. As the selection information of the coding mode group, the% can realize a decoding device that can implement a good decoding rate. Regarding the image decoding device of the present invention, a plurality of coding mode groups are registered with the coding mode group corresponding to the low bit rate registered by the coding mode corresponding to the bit rate of the original JMi · Gensan A , And a bit higher than the above-mentioned reference bit rate -----------_________. -......... The high bit rate corresponding to the encoding mode registered in the encoding mode group 10 This paper standard applies to China Standard (CNS) Α4 specification (210X297 mm) (Please read the precautions on the back before filling this page) * 1T Printed by the Consumer Standards Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs A7 B7 5. Composition of the invention (8). ______ In this way, an image decoding device capable of supporting a wide range of bit rates from a low bit rate to a high bit rate can be realized. Brief Description of the Drawings Fig. 1 is a drawing showing a structure of a previously encoded bit stream that generally transmits and receives signals between an image encoding device and an image decoding device. > / FIG. 2 is a diagram showing an MBTYPE table (Mb type-〇) for B-VOP encoding of VM8.0. Fig. 3 is a block diagram showing the structure of an image coding device according to the first embodiment of the present invention. Fig. 4 is a block diagram showing the structure of a coding mode determination unit according to the first embodiment of the present invention. Fig. 5 is a diagram showing a structure of a coded bit stream according to the first embodiment of the present invention. Fig. 6 is a flowchart showing the operation of the image encoding device according to the first embodiment of the present invention. Fig. 7 is a block diagram showing the structure of an image coding device according to a second embodiment of the present invention. 'FIG. 8 is a block diagram showing the configuration of the MBTY-PE determination unit according to the second embodiment of the present invention. Fig. 9 is a diagram showing the internal guest v of the MBTYPE-1 table in the second embodiment of the present invention. Fig. 10 is a diagram showing the contents of the MBTYPE-2 table in Embodiment 2 of the present invention. 11 (Please read the notes on the back before filling this page)-Order! ^ Wi. This paper size applies to Chinese national standards (CNS> A4 specification (210X297 mm)> Printed by the Consumers Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 5 Explanation of the invention (9) FIG. 11 is a flowchart showing the operation of the image encoding device in the second embodiment of the present invention. FIG. 12 is a diagram showing the structure of the bit stream of the winter knitting in the second embodiment of the present invention .. .... ~ ------ Schematic. Figure 13 is the second drawing of MBTYFE-3—table's hemp capacity. Figure 14 is the drawing of MBTYPE-leather table __ 's tolerance. .....— -....... ^^ Figure 15 shows the U of the contents of the MBTYPE-5 table. Figure 16 shows the contents of the MBTΎPEE-6 table. Figure 17 The figure is a diagram illustrating the modification of the direct prediction. Fig. 18 is a diagram showing the contents of the MBTYPE-7 table. Fig. 19 is a block diagram showing the structure of an image encoding device according to Embodiment 3 of the present invention. Fig. 20 FIG. 21 is a block diagram showing the configuration of the MBTYPE determination unit in Embodiment 3 of the present invention. FIG. 21 is a flowchart showing the operation of the pre-quantization process in Embodiment 3 of the present invention. FIG. 22 shows the present invention. The structure of the wide-range element stream in the embodiment; _scheme. Figure 23 is a block diagram showing the structure of the image code device of the fourth embodiment of the present invention. Figure 24 is a diagram showing the fourth embodiment of the present invention. Block diagram of the structure of the variable-length decoding section of the syntax solution board a. Figure 25 is a block diagram showing the structure of the change compensation section of the embodiment 4 of the present invention. 12 This paper size applies the Chinese National Standard (CNS) A4 specification (210X 297mm) (Please read the notes on the back before filling out this page) Thread-m- · Printed by A7 B7, Shellfish Consumer Cooperative, Central Bureau of Standards, Ministry of Economic Affairs 5. Description of the invention (10) Figure 26 is a drawing Shows the shadow of the fourth embodiment of the present invention—the operation of the image interpretation device ....... ~ The flow chart. ^ No. 2.7—Picture J 糸 —Yuanfa—Mingshi-Shiliezhi ~ Sentence-Method-Solution -Analysis, ... / Variable Flowchart of Variable Length Decoding Unit. Figure 28 shows the flow chart of the operation of the change compensation unit in the fourth embodiment of the sound. Figure 29 shows the flowchart. The block diagram of the structure of the image decoding device in Embodiment 5 of the present invention. FIG. 30 is a diagram illustrating the ^ syntax analysis of the embodiment 5 of the present invention. Block diagram of the structure of the decoding unit. FIG. 31 is a block diagram showing the structure of the change compensation unit in Embodiment 5 of the present invention. FIG. 32 is a flowchart showing the operation of the decoding device in Embodiment 5 of the present invention. Fig. 33 is a flowchart showing the syntax of the fifth embodiment of the present invention. 5]: Flow chart of the operation of the decoding unit. Fig. 34 is a diagram showing the decoding of the MBTPE-5 table according to the fifth embodiment of the present invention. Flowchart of sequence I. FIG. 35 is a flowchart of decoding sequence II of the MRTYPE-6 table according to the embodiment r5 of the present invention. Fig. 36 is a flowchart showing the operation of the change compensation unit in the fifth embodiment of the present invention. Fig. 37 is a block diagram showing the structure of an image decoding device according to Example 6 of the present invention. 13 This paper size applies to China National Standard (CNS) Α4 size (210X 297 mm) --------- Γ ----_---. 玎 ------ (Please read the back first Please note this page and fill in this page again) A7 V. Description of the invention (11) Figure 38 shows the structure of the decoding section. The second example of the structure of the decoding section is the red peak solution, the ear changes the length. Christine's flowchart. The solution of the sixth embodiment of the sound month is-installation-placement of the moving body 40. The diagram shows the flow of the operation of the decoding unit. ^ The analysis of the sixth embodiment and the description of the best embodiment of the variable length. The preferred embodiment will be described in detail. /% Description of the first embodiment of the present invention in accordance with the accompanying drawings The feature of the shadow silly cesium rice plate device of the first embodiment from the coded image area is a plurality of corresponding hi-hats Y-'s on the gray corresponding to the previous choreographer. In the coding mode, select the most suitable image to be edited ~ u The specified patterns on the group are one by one, and the image is more windy. Use the table to turn— 槔 may be the distinguishing code * ~ _., One ~~ one With regard to the image editing device of the present invention, the coding of each phase image of a shirt image sequence constituting a digital dynamic image is implemented as a unit. Since each 'y edge is owed to a $ code into a moving edge region, the minimum coding unit is The image area is called, the coded image area, and the image of each time formed by the collection of the coded image area is called "the image". The example of the coded image area is IS0 / IEC Examples of the microblock 'encoded image explicitly indicated by 13818_2 are the image frame of the video signal, the video object plane (VOP) clearly indicated by IS0 / IETC JTC1 / SC29 / WG11 N1796. This paper scale is applicable to the country of China 檩Standard (CNS) A4 (210X297 mm) (Please read the notes on the back before filling in this page> -Order

-M Printed by the Consumer Cooperatives of the Central Bureau of Accreditation of the Ministry of Economic Affairs A7 B7 V. Description of the Invention (12) Figure 3 is a block diagram showing the structure of the image coding device in Embodiment 1 of the present invention. In the figure, 1 is a subtraction unit, 2 is a coding mode determination unit, 3 is an orthogonal transform unit, 4 is a quantization unit, 5 is a variable-length encoding / multiplexing unit, 6 is an inverse quantization unit, and 7 is an inverse quantization unit. The orthogonal transform unit, 8 is an addition unit, 9 is a memory, 10 is a change detection unit, 11 is a change compensation unit, and 12 is a conversion unit. Fig. 4 is a block diagram showing the configuration of the encoding mode determination section 2 shown in Fig. 3. In the figure, 21 is a determination processing unit, 22 is a coding mode table A, 23 is a coding mode table B, and 24 is a conversion unit. The coding mode determination unit 2 includes two coding mode tables A22 and B23 including identifiers or identification numbers of coding modes that can be obtained from i-coded images such as ciphers that are variable-length-encoded. ..... -... "In the first embodiment, on the one hand, the encoding mode table ^: 2 includes the encoding mode table used when encoding at a bit rate lower than a predetermined reference bit rate- -..---------- | | _ The encoding mode group formed by the complex encoding mode as shown in the previous Figure 2. On the other hand, the encoding mode table _ on 23 series contains the image overlay Group of coding modes when the rabbit is coding at a bit rate higher than a predetermined reference bit rate. '..... Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs (please read the precautions on the back before filling out this page) Figure 5 shows the encoding output of the image encoding device according to Embodiment 1 of the present invention A fabricated pattern of the encoded bit stream of the signal. This coded bit stream 213 is the same as the previous coded bit stream _ shown in FIG. 1, because each time a frame (1 day surface) is created, it is encoded by the coded image

The I header information is composed of each coded data of each region 1 to N when one frame is divided into, for example, N regions. However, unlike the previous coded bit stream shown in Figure 1, it is printed on 15 paper sizes to the Chinese National Standard (CNS) A4 (210'〆297 mm). It is printed by the Consumers' Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs. A7 B7 V. Description of the invention (13) _ The encoding mode table selection information 200 bit information is newly added as the encoding mode group selection information to the transcoded image header information. One Li Yuan is only enough for two encoding mode tables A22 and B23. When setting more encoding modes .................... The number of 200 bits must increase with the number of selectable tables. Each of the encoded data of each region 1 to N is the same as the previous encoded bit stream shown in FIG. 1. The encoded data is composed of compressed image data 205 and, for example, encoding mode information 206 corresponding to each quantization mode. The quantized step size, the information 207 and the change information 208 are composed of overhead information. Next, the operation will be described. Fig. 6 is a flowchart showing the operation of the image encoding device according to the first embodiment of the present invention. (1) Selection and decision of encoding mode. First, in the first step ST1, when this encoding device is activated, the user selects the use of switches and programs according to the conditions such as the bit rate used by the device. The coding mode table selection information 200 of the coding mode tables A22 and B23 is output to the coding mode determination unit 2 and the variable-length coding / multiplexing unit 5. In short, when the user encounters the driving of the encoding device, the target bit rate can be set. The set bit rate has a threshold to distinguish the low bit rate from the high bit rate. This is the encoding mode table selection information 200. Next, in step ST2, the conversion unit 24 of the encoding mode determination unit 2 selects the information 200 based on the encoding mode table. When encoding at a low bit rate, it selects and determines the encoding mode table A22 corresponding to the low bit rate, and encodes at the high bit rate. At this time, the encoding mode table B23 corresponding to the high bit rate is determined. 16 (Please read the precautions on the back before filling this page) Γ 线 线-#-This paper size is applicable to China National Standard (CNS) A4 (210X297 mm) Printed by A7 B7, Employees ’Cooperative of Zhongjia Standard Bureau, Ministry of Economic Affairs V. Description of the invention (14) (2) Generation of coded video signal and selection of coding mode Next, the determination processing unit 21 of the coding mode determination unit 2 with respect to the input video signal 201 of the input coded video area, Using the processing of the following steps ST3 to ST14, the encoded video signal 5 corresponding to all the available encoding kernels is selected to select the encoding mode with the best encoding efficiency. The encoding mode is described here. The mode in which the input image signal is used as the encoded image signal is also an encoding mode. Usually, it is switched to the internal mode. In addition, the prediction shadow percentage is generated by using the change compensation prediction, and the obtained prediction error letter @ _ $ is the encoded video signal, and the mutual mode is used. In mutual mode, the prediction mode is used to set the accompanying encoding mode. For example, for B-pictures not specified in ISO / IEC 13818-2, change-prediction prediction is performed from previously completed coding images, and change-compensated prediction is performed from previously completed coding images. In the subsequent direction prediction, the prediction images obtained through the forward direction prediction and the backward direction prediction are added and averaged as prediction images, and a plurality of prediction methods such as two-direction prediction can be obtained, and various coding modes are set correspondingly. From such a plurality of encoding modes, selecting a one-to-one r .—- -........_ ** ** The coded video signal obtained from the mode with the best code efficiency becomes compression coding. Object. Therefore, in the coding mode of the internal mode, although the input video signal 201 is used as such a coded video signal, it does not need to be generated. In the coding mode of the mutual mode, prediction of various prediction modes corresponding to the possible selection of the mutual mode The error signal is generated using the change detection section 10, the change compensation section 11, and the subtraction section 1 of FIG. First, in step ST3, the change detection unit 10 applies the Chinese paper standard (CNS) A4 specification (210X297 mm) from the 17 paper sizes in the memory 9. (Please read the precautions on the back before filling this page) Order Line A7 A7 Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the invention ¢ 5) The reference image signal 211 obtains the change information 208 corresponding to the prediction method. Next, in step ST4, the change compensation unit 11 reads the reference image 211 from the memory 9 based on the change information 208, and performs necessary calculations to generate a prediction image 212. In this way, the prediction video 212 is subtracted from the input video signal 201 in the subtraction unit 1 to obtain a prediction error signal 202 or a coded video signal 203. The above processing is performed by all the encoding modes registered in the selected encoding mode table. Then, in step ST5, it is checked whether or not the processing for generating the encoded video signal is completed for all the encoding modes registered in the selected encoding table. When it is finished, in the next step ST6, the determination processing unit 21 evaluates and determines the encoded video signal 203 corresponding to each encoding mode included in the selected encoding mode table, and selects the best encoding efficiency at the same time. For the encoded video signal 203, the encoding mode corresponding to this is selected as the encoding mode of the encoded image area, and the encoded video signal 203 and the identification factor are output as the encoding mode information 206, respectively. (1) Compression encoding processing In the next step ST7, the orthogonal transform unit 3 converts the encoded video signal 203 output from the encoding mode determination unit 2 into an orthogonal transform coefficient 204 that performs a transform such as DCT (Discrete Cosine Transform). . Then, in the subsequent step ST8, the quantization unit 4 quantizes the orthogonal transform coefficient 204 in a predetermined quantization step, outputs the compressed image data 205, and outputs a quantization step indicating the quantization step. Information 207. 18 This paper size applies to China National Standard (CNS) A4 specification (2 丨 0X297 mm) (Please read the notes on the back before filling out this page) • iv Order printed by the Consumer Standards Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 5. Description of the invention (16) (2) Variable length encoding and multiplexing In step ST9, the compressed image data 205, encoding mode table selection information 200, encoding mode information 206 of the encoding mode determination unit 2 and changes in the quantization unit 4 The additional information such as the change information 208 of the detection unit 10 is converted into a bit string using variable-length coding and the multiplexing unit 5 is multiplexed in accordance with a predetermined syntax. As shown in FIG. 5, a 1-bit code The mode table selection information 200 itself is set as encoded header information, which is output as the encoded bit stream 213 ′ of the structure. The quantization step information 207 indicating the quantization step size of the quantization section 4 may be known and registered in advance on the variable-length coding / multiplexing section 5 side. (3) The local decoding process compresses the video data. In steps ST10 and ST11, the inverse quantization unit 6 and the inverse orthogonal transform unit 7 are used to obtain a locally decoded prediction error signal that is locally decoded into a coded video signal. 209. Then, in step ST12, the conversion unit 12 sends the necessary prediction image 212 to the addition unit 8 according to the encoding mode information 206, and the addition unit 8 adds the locally decoded prediction error signal 209 and the prediction image 212 to obtain a locally decoded image. Signal 210. Next, in step ST13, the local decoded video signal 210 is stored in the memory 9 for subsequent encoding. In the last step ST14, for the continuous processing of steps ST3-ST13, it is checked whether all the coded image areas have ended, and when all the coded image areas have ended, the above coding process is ended. As above, if according to this embodiment 1, corresponding to the target bit rate and other conditions, 19 (Please read the precautions on the back before filling out this page) The paper size of this edition applies the Chinese National Standard (CNS) A4 specification (210X297 mm) Printed by A7 B " 7 of the Consumers 'Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the Invention (17) First prepare each pre-assigned coding pattern table as a plurality of suitable coding pattern groups' 5 corresponding conditions when this device is started Set the encoding mode table selection information, and select the appropriate encoding mode table, and at the same time select the most suitable encoding mode from the selected encoding mode table. The encoding mode table selection information is not in every encoding data in each region. Since it is set to be the encoded image header information, by setting a plurality of encoding mode tables, although the encoding mode is increased, the amount of encoding mode information in the additional information of the encoded data has not increased, and a good effect can be obtained. Encoded image encoding device. To be more specific, for example, at a high bit rate, because the entire amount of information in the encoding bit $ is sufficient, if there are many 殁 code patterns-and-can be adapted to 迤 encoding, even if the encoding mode information is added; There are many words, and overall, the coding efficiency can be improved. However, at the low bit rate I, because I have to suppress. __________________ ----------------------------------- -------- " s ---------------------- The information of Panyi whole farmer is halo.. Xiangli full cover 敫 like farmer The share of surname pool information, including additional information such as coding mode information, has a greater proportion. Therefore, if a large number of encoding modes are used, the exhaustion efficiency of compressed image data will not be improved, which will cause the burden of additional information such as the encoding mode information m to increase, which will cause the encoding efficiency to deteriorate. For a coding device that has only one coding mode table fixed, when the bit rate is changed, it is difficult to maintain adaptability even if the coding efficiency does not deteriorate. For the present embodiment, although an example of conversion of a coding mode table corresponding to a communication condition of a bit rate is shown, it does not limit the present invention. For other conditions, for example, a time resolution corresponding to an image (such as a frame rate) , Encoding size of the encoded image (using the length and width of the video object plane), etc. 20 This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) --------- C --- ------- tr ------- If (Please read the notes on the back before filling out this page) A7 B7 V. Description of the invention (18) Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs Can be converted. In short, if the frame rate is different, the encoding image can improve the encoding efficiency due to different prediction methods. Also, — \ 一 ® Yu Shier'er, “The size of a broken coded image represents the empty and two M freshness of the moving image sequence of the coding object.” This is a quantity that directly reflects the target bit rate. + π does not change the same encoding mode table as the example. This is also true for the other embodiments, and for this embodiment i, explaining that the two encoding modes corresponding to the low bit rate and the high bit rate are provided with respective encoding mode tables A and B. For the invention, corresponding to the bit rate Even if there are many H number of equal-edged exhaustion pattern groups, it is sufficient. X, it is also possible to register a plurality of coding mode groups on the i coding mode tables. At this time, the conversion unit of the 'coding mode determination unit 2 selects information using the coding mode table, and selects i from the plurality of coding mode groups. Coding pattern groups and transforming them. This is the same as the other embodiments. In the first embodiment, it is described that the encoding mode table selection information is set to the encoded video header information on the encoded video such as video, etc. For the present invention, 'is not limited to the encoded video'. On the specified encoded image, the encoded image header information is set as the encoded image header information. When new encoded image header information appears, it is based on the first detected encoded image header information. Encoding mode table selection information 'Select encoding mode table' or relative to all encoded images in the upper hierarchy as the encoded image header information, setting the encoding mode table selection information to header information can also For encoding video (such as MPEG-4 VOL header information, etc.), select one encoding mode table. This is the same for other embodiments. 21 This paper size applies to Chinese National Standard (CNS) A4 (210X297 mm) (Please read the precautions on the back before filling this page)

r'Y order-Φ- · A7 B7 V. Description of the invention (19) For this embodiment 1, compensation is performed on all coding modes registered in the selected one's coding table to 'change' to obtain a prediction error signal. Generate an encoded video signal, evaluate the encoded video signal, and select the encoding mode that is determined to be the most efficient. In the present invention, such evaluation and determination are not performed, and other methods may be used to select the most efficient encoding mode from a plurality of encoding modes. Good coding mode. Embodiment 2 The video encoding device according to the second embodiment is a video encoding device according to the first embodiment shown in Fig. 3, which is improved in accordance with the MPEGZ4 standard. — 一 '… ----- ...... Fig. 7 is a block diagram showing the structure of the image coding device of the second embodiment. In the figure, 3 1 series MBTYPE (microblock coding type) determination section, 32 series DCT section, 33 series inverse DCT section, 34 series shape coding section, and others with the same reference number are the same as those in the third figure of the first embodiment. The configuration shown is the same. The difference from the encoding device of the first embodiment shown in FIG. 3 is described. In the second embodiment, an MBTYPE determination unit 31 serving as the encoding mode determination unit 2 and a DCT unit serving as the orthogonal transformation unit 3 are provided. 32. Do-〆 __________------------------is the inverse DCT section 33 of the inverse orthogonal transform section 7, and the other new settings are ... ----- ^ --- -~ ------........ ........ Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs ---------- r ''. Ϊ --(Please read the precautions on the back before filling out this page> MPEG-4's unique shape coding section 34, the compressed shape data 301 is variable length coding • multiplexing section 5 performs variable length coding • multiplexing The local decoded shape data 303 of the shape coding unit 34 is input to the MBTYPE determination unit 31, the DCT unit 32, the inverse DCT unit 33, the change compensation unit 11, and the change detection unit 10. However, even MPEG-4 When a rectangular image frame such as a video signal is used as the encoded image, 22 paper sizes are applicable to the Chinese National Standard (CNS) A4 specification (210 × 297 mm). The Consumers Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs has printed A7. B7 V. Description of the invention (20) Since the shape data must be rectangular, no shape coding unit 34 is needed at this time. Figure 8 shows the structure of the MBTYPE determination unit 31 of Figure 7. Block diagram. In the figure, 41 is a decision processing unit, 42 is an MBTYPE-1 (microblock encoding type 1) table, 4 is an MBTYPE-2 (microblock encoding type 2) table, and 44 is based on the MBTYPE table " Reconstructive Information- Zhuan Yanzhuan: Change ~ _ Department.

FIG. 9 is a drawing showing the contents of MBTYPE-1 table 42 in FIG. 8. In the figure, on this MBTYPE-1 table 42, the forward prediction of pattern numbers 0 to 3 (the value of MBTYPE), and after! To Pre-Li + quantization conversion, forward direction prediction + ..... ~ 4 獐 cans of mamowu for conversion and stuffing, each of which is recorded, and the X that is used in encoding is recorded through its own encoding mode The quantized step size difference (DQUANT), forward direction change vector difference value (MVDf), backward direction change vector difference value (MVDb), direct prediction difference vector (MVDB), and representation Secrets of the respective coding modes β This MBTYPE-1 table 42 is a table obtained by adding the padding to match the bit rate to the two-direction prediction + quantization conversion mode instead of the MBTYPE-O table shown in Figure 2 " The padding is an encoding of dummy bits to match the bit rate of the bit rate. When decoding padding equivalent to padding bits, etc., the padding cipher is discarded during decoding. Then, the subsequent microblock data will not exist. In this way, the bit rate can be incorporated in small microblock units. Therefore, as shown in Table 42 of MBTYPE-1 in the coding mode of padding, not all data are used. The cipher words of each coding mode are 23 paper sizes. The Chinese National Standard (CNS) A4 specification (210 X 297 mm) is applicable. (Please read the ir notice on the back before filling out this page} ----.- IT- -----. ilI® ------ A7 B7 V. Description of the invention (21) It is determined according to the occurrence accuracy rate of each coding mode, and the best secret word is not decided at present. Figure 10 is a drawing Figure 8 shows the contents of MBTYPE-2 table 43. In the figure, MBTYPE-2 table 43 shows a 9-coding mode with mode numbers 0 to 8 _, which is internal coding, coding, and internal Encoding + Quantity: Sub-transformation, _______________-. .... ---------- Two-way prediction, two-way prediction + quantization conversion, backward direction prediction, post -... ..... .-.. Direction prediction + quantization conversion, forward direction prediction, forward direction prediction + quantization conversion and padding, each of them is registered in the quantization step size of the data to be used for encoding through the respective encoding mode Difference value (DQUANT), change vector difference value for forward prediction (MVDf), change vector difference value for backward prediction (MVDb), and their respective coding modes Secret words. Next, the operation will be described. FIG. 11 is a flowchart showing the operation of the image encoding device according to the second embodiment of the present invention. Here, the explanation will focus on the differences from the operation flow of the first embodiment. Step ST22-1 is removed, steps The processing of ST21 to ST34 is equivalent to the processing of steps ST1 to ST14 in FIG. 6 of Embodiment 1, except that the name of the processing is different, and the processing is substantially the same as that of the corresponding step. Printed by a consumer cooperative (please read the precautions on the back before filling out this page) For Example 2, between steps ST22 and ST23, add step ST22-1, that is, use the shape coding unit 34 to perform shape coding processing in microblock units. Through this processing, the compressed shape data 301, the locally decoded shape data 303, and the like are output. FIG. 12 is a diagram showing the structure of the encoded bit stream 3 16 encoded and output by the image encoding device of Embodiment 2. Here, VOL (Video Object 24 This paper size is applicable to Chinese National Standard (CNS) A4 specification (210X297 mm)) Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (22 )

Layer) is a component of a moving image object v〇 (Video Object) that includes a time axis, and is composed of a plurality of 乂 〇? ('^ 丨 (160〇Ca (^? 1 & 1 ^). \ ^ 0? Department Shows the status of V0 at each moment. Describes the difference from the encoded bit stream of the first embodiment shown in FIG. 5. In this second embodiment, 'because it is specifically MPEG-4, the encoded video information is VOP standard. The header information 'in the VOP header information, or in the VOL header information of the video object layer of the higher If layer' is set as the encoding mode group selection information, and the MBTPEE table selection information (b VOP mode type) 302. The compressed shape data 301 is also set as additional information of the encoded data of each of the encoded image regions 1 to N. 的 The method of setting the MBTYPE table selection information (B VOP mode type) 302 in the VOL! Header information is greater than that in each VOP Once the VOP header information is set, the MBTYF6 table selection information 302 is set, and the MBTYPE table meal—data tfLj_〇2 can reduce the amount of nitrogen and increase the coding efficiency. At this time, it is not possible to select each coding mode table of the VOP. 》 Encoding side information (MBTYPE) 308 in the additional information The secret of every choreography mode of MBTYgE is set. On the one hand, the quantization step size information in the additional information is 309....... Is set when using DQUANT. For example, that is, The case of the MBTYPE-1 table in FIG. 9 is the case of the encoding mode of backward prediction + quantization conversion, and the case of the encoding mode of forward prediction + quantization conversion. The dynamic information in the additional information uses MVDf, MVDb and MVDB are set. As described above, 'If according to Embodiment 2 which is specifically MPEG-4, relative to 25 ---------- t, ------ IT ------ ^ 1®, I (Please read the precautions on the back before filling this page) Exhaust.. Standards of Standards (CNS) A4 specifications (2H) X297 mm A7 A7 Employee Consumer Cooperatives, Central Standards Bureau, Ministry of Economic Affairs Printed ___B7 _____ V. Description of the invention (23) The first embodiment is actually added; the shape editor belongs to -3 inches. The compressed shape data is 301 spine variable length coding and multiplexing, and the bureau | decoded shape data is in DCT. , Inverse DCT, change detection and change compensation are used. The selection of the MBTYPE table of the encoding mode table is the same as in the first embodiment. The situation is the same as in the case of Embodiment 1. Even if the encoding mode is increased by setting a plurality of encoding mode tables, the amount of encoding mode information 308 in the additional information of the encoded data is not increased, so that efficiency can be obtained. Good winter coding video coding equipment Qin For the second embodiment, by! The padding encoding mode is recorded on both MBTYPE tables 42 and 43. Compared with the case where there is no padding encoding mode, the bit rate can be merged into the bit rate β-... ≪ ',. ≫ The MBTΜPEE-2 table 43 in Fig. 10 not only for padding, but also the direct prediction of the proposed MBTΥPEE-0 table shown in Fig. 2 is changed to internal encoding. The quantization step size of each coding mode is switched to ON / OFF mode, where each coding mode corresponds to each coding mode of internal coding, two-direction prediction, backward direction prediction, and forward direction prediction. In addition, even when encoding in any mode, the quantization step size can be changed, and if the quantization step value of a nearby microblock does not change, it is not necessary to provide quantization step information (DQUANT ) 309 is multiplexed into a bit stream, and a code 0 with good efficiency can be obtained. For Embodiment 2, the MBTYPE table of the encoding mode table is taken as an example to illustrate MBTYPE-1 shown in FIG. 9 and FIG. 10 Table and MBTYPE-2 table, for the purposes of the present invention, It is not limited to these tables, for example, to make 26 paper sizes apply the Chinese National Standard (CNS) A4 specification (210X297 mm) --------- r-^-T.. * ' -(Please read the precautions on the back before filling this page) Printed bags of the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economics of the Thread ΑΊ ____ Β7 V. Description of the Invention (24) Use the code to remove the padding from the MBTYPE-1 and MBTYPE_2 tables A table of modes, or a coded mode that removes padding from the MBTYPE-2 table, or a table of modes in which quantization step size conversion 0N / 〇FF is prepared on both sides, where quantization step size conversion 0N / The OFF mode corresponds to the encoding of the internal encoding, internal encoding + quantization conversion on the table with direct prediction, direct prediction + direct prediction of quantization conversion, two-direction prediction, backward prediction, and forward prediction. mode. For the encoding mode table, for example, the MBTYPE-3 table shown in FIG. 13 and the MBTγρΕ_4 table shown in FIG. 14 may be used. If this is the case with the above-mentioned embodiments 1 and 2, since there are multiple encoding modes, even if the encoding mode is increased, the amount of additional information of the encoded data will not increase, and the amount of additional encoding mode information will not increase. The impact of the conditions T .., can still be implemented-the efficiency I is good., Edit-, two have the same prediction effect for direct prediction and two-direction prediction 'via the VOL header information or ν〇P header The MBTQPE table selection information (BV〇P m〇de type) in the information is adaptively converted according to the coding bit rate, the direct prediction of the change vector encoding amount is not much, and the hope is to improve the prediction efficiency in a wider exploration range. Direction prediction 'Even if the quantization conversion is added to all modes, the number of bits representing the secret of MBTQPE can be saved. Alternatively, the MBTγρE_2 table shown in Fig. 10 and the MBTQPEE-5 table shown in Fig. 15 may be used. In this way, as in the above case, since the "encoding i-mo-text table has a plurality of numbers", even if the encoding mode is increased, the amount of encoding mode information in the additional data sfl of the encoded data will not increase. It is still possible to implement efficient editing. In particular, MBTYPE-5 --------- Γ, — (Please read the notes on the back before filling out this page) -β 线-#-r This paper size applies to China National Standard (CNS) A4 (210X297 mm) A7 B7 printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs 5. Description of the invention (25) Tables are added to the MBTYPE-3 table shown in Figure 13 to match the bit rate. Filled table. When the equivalent cipher is decoded, the subsequent microblock data does not exist. Therefore, a small bit rate according to a microblock unit can be met. The filled cipher is only read and discarded when decoding. Alternatively, the MBTYPE-2 table shown in Fig. 10 and the MBTYPE-6 table shown in Fig. 16 may be used. In this way, similarly, even if the amount of coding mode information in the additional information of the coded data is not increased, efficient coding can be performed under the influence of the given conditions. In particular, because the MBTYPE-6 table uses a modified direct prediction mode that does not use a direct prediction difference vector (MVDB) as the table of the direct prediction mode of the MBTYPE-1 table shown in FIG. 9, the direct prediction is performed by using this modification. In the microblock decoding mode, the decoding process of the direct prediction differential vector (MVDB) becomes incomplete. As the amount of additional information is reduced, the coding efficiency is improved accordingly. Modified straight | prediction mode, as shown in Figure 17, unlike direct prediction, the direct prerequisite jJ difference vector (MVDB) is usually 0, which means that the direct prediction difference vector (MVDB) is not used, which uses the change vector Itself, _______--------------- .. Find the f-side H service change to halo (MVf), change in the backward direction prediction_also * (MVb). Alternatively, the MBTYPE-2 table shown in Fig. 10 and the MBTYPE-7 table shown in Fig. 18 may be used. In this way, even if the encoding mode is increased, the amount of encoding mode information 308 in the additional information of the encoded data will not increase, and efficient encoding can still be performed under the influence of the given conditions. In particular, because the MBTYPE-7 table is the 28 shown in Figure 15 (please read the precautions on the back before filling this page) r'1 Order-#-· This paper size applies to China National Standard (CNS) A4 Specifications (210X297 mm) Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (26) ~~ The direct prediction mode of the MBTYPE-5 table was changed to the modified direct prediction mode without using the differential vector (MVDB) Table, from the Huangzheng direct prediction model, in the process of applying fresh codes to the gas area, the process of directly predicting the difference vector (MVDB) is not completely necessary. As the amount of additional information is reduced, the coding efficiency is improved accordingly. —, '. Embodiment 3 19th | #' It is not a block diagram of the structure of the image coding apparatus according to Embodiment 3. In the picture, 51 is based on the chapter pre-chemical method to select the information 402 two actions! 1ίΒΤ 翌 ft, 52 series are selected according to the quantization method. Conversion section for conversion, 53 series H.263 type quantum north-part _, meaning ^] VJPEG-2 type 1 sub-chemical part, 55 series h.263 The type inverse quantization unit and the MPEG-2 type inverse quantization unit are those having the same reference numbers as those having the same configuration as shown in FIG. 7 of Embodiment 2. In short, the video encoding device of the third embodiment is described in more detail in the structure of the quantization unit of the video encoding device of the second embodiment corresponding to the MPEG-4 standard shown in FIG. Corresponding to the use of low .—.....—… .......... ^^ ·% f rate of H.2 ^ 3, lattice H · 263 type quantization unit 53 and jj · 263 type 'inverse quantum The quantization unit 55 ′ corresponds to the MPEG-2 using a high bit rate, the MPEG-2 type quantization unit 54 and the MPEG-2 type inverse quantization unit 56, and the The quantization method selects a video object layer quant type 402 and converts the conversion unit 52. Fig. 20 is a block diagram showing the configuration of the MBTYPE determination unit 51 of the third embodiment shown in Fig. 19. MBTYPE judging unit 51 of the image coding device of the third embodiment is based on the paper size of China National Standard (CNS) A4 (210X297 mm) (please read the precautions on the back before filling this page), -'0 line

Printed by the Consumer Standards Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the Invention (27) The MBTYPE determination section 31 of the second embodiment shown in FIG. 8 is the same, except that, as shown in the figure, the conversion section 61, replaced MBTYPE_ ..... _ ------- ....... table selection information 302, and converted MBTYPE-1 table 42 and MBTYPE-2 based on existing quantization selection information 402 Table 43. The other structures are the same as those of the video encoding device of the second embodiment, and therefore are denoted by the same reference numerals, and descriptions thereof are omitted. Next, the operation will be described. The operation of the third embodiment is basically the same as that of the second embodiment shown in FIG. However, in Embodiment 3, the quantization method is performed by using the H.263 type quantization unit 53 and the MPEG-2 type quantization unit 54. ---------------------- ----------- ........ .............................- conversion, so the following processing is different. Fig. 21 is a flowchart showing the operation of the quantization process in step ST28 of Fig. 11. In step ST41 in FIG. 21, the conversion unit 52 in FIG. 19 selects information 402 based on the existing quantization method, and determines whether the quantization method is the H.263 method. Here, when the quantization method selection information 402 is H.263, in the next step ST42, the H_263 type quantization unit 53 performs quantization according to the MBTYPE-5 table corresponding to the low bit rate. On the other hand, when the quantization method selection information 402 is MPEG-2, in step ST43, the MPEG-2 type quantization unit 54 performs quantization according to the MBTYPE-2 table corresponding to the high bit rate. In the third embodiment, in the inverse quantization process in step ST30 in FIG. 11, the conversion unit 52 is in the step ST41 in FIG. 21 when the quantization method selection information 402 is determined to be H.263, and the H.263 type Department of Inverse Quantization 55 30 This paper size applies Chinese National Standard (CNS) A · 4 specifications (210X 297mm) I -------- C ------, βτ ------ # -111- (Please read the notes on the back before filling this page) Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs A'7 —------------ B7 V. Invention Description (28 ) A 'Perform inverse quantization according to the MBTγρΕ_ 丨 table corresponding to the low bit rate. On the other hand, when the quantization method selection information 402 is judged to be MPEg_2, the MPEG-2 type inverse quantization unit 56 performs inverse quantization according to the MBTYPE-2 table corresponding to the high bit rate. Fig. 22 is a diagram showing the structure of the I-bit stream of the coded output of the image conversion device of the third embodiment. The difference from the encoded bit stream of the second embodiment shown in FIG. 12 will be described. In the third embodiment, the existing 夂 quantization method selection information (video 〇bject iayer quant) which stores the v0L header information is used. type) 402, MBTYPE table selection information (B VOP móde type) 302 as the coding mode group selection information, and does not reset the MBTYPE table selection information (B VOP móde type) 302 to VOL header information or VOP Header information. Therefore, as long as the MBTYPE table selection information (B VOP modetype) 302 is not reset, the amount of VOL header information or VOP header information of the encoded bit stream 416 + will not increase, without any additional information being added to the encoding bit. In the current syntax of stream 416, encoding efficiency can be improved. As described above, "if according to Embodiment 3, since it is the same as the device of Embodiment 2 except for the conversion of the quantization method, it is the same as the case of Embodiment 2" even if the encoding mode is increased, the encoding mode information in the additional information of the encoded data The quantity will not increase, and under the influence of the given conditions, efficient encoding can still be implemented. It is taught by the teacher_using the existing quantum pool method selection information 4.02 in the YOL-header information as the MBTYPE table selection Zen information 302, none. One ^... ---... In the current syntax of the encoding bit paralysis 416, it is necessary to add a color image to an image encoding device that can improve encoding efficiency. ....... • 一 '31 This paper size applies Chinese National Standard (CNS) Α4 specification (210X297 mm) --------- r -------- IT --- --- (Please read the precautions on the back before filling this page) A7 B7 Printed by the Consumer Cooperatives of the Central Office of the Ministry of Economic Affairs V. Invention Description (29) For the H 263 specification using a low bit rate, the basis corresponds to the use The high bit rate MPEG-2 bit rate MPEG-2 can realize an image coding device that can correspond to a wide range of bit rates from low bit rates to high bit rates. For the embodiment 3, the mbtypej table and the MBTγρE_2 table are taken as examples. As described in the above embodiment 2, in addition, the MBT × PEE table corresponding to a low bit rate, such as MBTγρE 0, μΒΤΥΡΕ_3, ΜΒΤΒΡΕ-5, ΜΒΤΒΡΕ-6, MBTQPE-7 can also use MBTYPp tables corresponding to high bit rates, such as μβτυρε_4. Embodiment 4 The image decoding device of Embodiment 4 decodes the image at each moment constituting a scene sequence such as a digital dynamic image as a unit, and each image is decoded in units of a knife-shaped image area, corresponding to The encoded image area on the encoding device side is referred to as the image area with the smallest decoding unit. The decoded image to the image area 'corresponds to the encoded image on the encoding device side. The image is called ,,, and the decoded image. The examples of de-triggering are the microblocks explicitly indicated by the 13818 collar, and the examples of decoded images are the video block of the video signal, and the video object plane clearly indicated by ISO / IETC JTC1 / SC29 / WG11 Ν1796. The image decoding device of the fourth embodiment includes a coding mode table that selects detailed information and specifies the coding scale used according to the coding mode determined by the decoded image area, compared to the previous decoding settings for decoding. " Set the encoding mode group on the township special table—the encoding mode for each county and decode it. 32. Paper weaving is applicable to China (please read the precautions on the back before filling this page) --- 1T m fit— (i Kn line ---. J-Lr: A7 B7 printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs of the People's Republic of China (5) Invention Description (30) structure, corresponding to the decoding device of the image coding device of the first embodiment described above. _ Fig. 23 is a block diagram showing the structure of an image decoding device that expands and reproduces a compression-encoded digital image according to the fourth embodiment. In the figure, 71 is a syntax analysis and variable-length encoding unit, and 72 is inverse quantization. Section, 73 series inverse orthogonal transform section, 74 series change compensation section, 75 series memory, 76 series conversion section, 77 series addition section. Figure 24 shows the syntax analysis and variable length coding section shown in Figure 23 The composition of the block group. In the figure, the 81 series Code mode table selection information decoding section, 82 series conversion section, 83 series coding mode table A decoding section, 84 series coding mode table B decoding section, 85 series quantization step size decoding section, 86 series change information decoding section, 87 series compression Image data decoding section. Figure 25 is a block diagram of the structure of the change compensation section 74 shown in Figure 23. Figure. In the figure, the 91 series conversion section, the 92 series prediction method 1 implementation section, and the 93 series prediction method 2 implementation section The reference numerals 94 and 94 are implementation units of the prediction mode 3. The operations are described below. The operation is described below. Fig. 26 is a flowchart showing the operation of the decoding process performed by the video decoding device of the fourth embodiment. First, in step ST51 In Fig. 23, the syntax analysis and variable length encoding unit 71 analyzes the encoded bit stream 213 constructed in Fig. 5 and divides it into compressed image data 205, quantization step information 207, and change information --- ---------------------------- ........- 208 and other encoded data and output. At this time, A count of the position information in the daytime plane of each predicted image area such as a microblock, etc., is output after totaling. For this syntax solution Analytical processing is usually carried out after checking out 33 paper sizes that are applicable to Chinese National Standard (CNS) A4 specifications (210X297 mm) (please read the precautions on the back before filling this page) Order the printed bags of the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (31) is the start code of the unit segmentation of each set of frames, frames, etc., and the decoding of this hierarchy is performed accordingly, and the syntax analysis processing here is only for microblocks and other The process of analyzing the stages of the data in the decoded image area is targeted. The encoding mode information 206 is used to decode the data of the ranks such as the decoded image above the decoded image area. -In step ST52, the inverse quantization unit 72 performs inverse quantization on the compressed image data 205 using the quantization step information 207, and outputs an orthogonal transform coefficient 215. In step ST53, the inverse orthogonal transform unit 73 performs inverse orthogonal transform on the orthogonal transform coefficient 215, and outputs a decoded prediction error signal 216. The orthogonal transform system, such as DCT (Discrete Cosine Transform), is the same as the user on the encoding side. In step ST54, when decoding the decoded video region coded in the mutual mode, the change compensation unit 74 performs syntax analysis. The variable-length coding unit 71 inputs the change information 208 decoded from the coded bit stream 213 and the predicted video. The position information 214 in the day and time of the area is used to compensate for the change, and the prediction image 219 is taken from the reference image 218 of the memory 75 and sent to the conversion unit 76 °. The inverse quantization processing and inverse orthogonal transform processing of steps ST52 and ST53 The change compensation process in step ST54 may be performed even if the order is not followed, and the reverse order may be performed simultaneously. In step ST55, the conversion unit 76 determines whether or not the encoding is performed in the mutual mode based on the encoding mode information 206 decoded from the encoded bit stream 213 by the syntax analysis / variable length encoding unit 71. Then, in the case of the decoded image area coded in the mutual mode, in step ST56, turn to 34 (please read the precautions on the back before filling this page) Γ · Order, line-UU · This paper is again suitable for China National Standard (CNS) A4 Specification (210X 297mm) No. 86115728 Revised Sheet Λ7 B? ^ Date: 88 · 1 · 26 V. Description of the Invention (32) Change Unit% Output prediction image 219 to Addition Unit 7 The prediction image 219 is added to the 7th inverse orthogonal transform section, and the addition section 77 is No. 216, and is output as an exhaustive image signal mirror 217. ^ Prediction error signal Decoded image region encoded by the internal mode. On the other hand, in the above, the conversion section 76 will rotate out. In the addition section 77, the second and lower decoded prediction error signals of the orthogonal transform section 73 in step ST57 are inverse signal 217. The output is decoded in the step ST58 of the Decoding Video Association. Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs: The code 71 is compiled according to the figure shown in Figure 5. Shadow ::: =: Etc. Every-decoding process 2 == all the encoded data in the case where the encoded data of the image area can be detected and there is also 'Dong Dong Fu from the initial step ST51 due to the decoding of the security guard U AL ^ case ^ * method Detected encoded video encoding data, etc. "Because all the decoded encoded data has not been processed for decoding. This decoded video signal 217 is a f: = ^ device (not _, a later Sr. loss image 218, and is written into the memory 75. Fig. 27 shows the syntax analysis and the operation of the variable-length encoding unit η. That is, the flowchart of the processing of step ST51 of the 26th objective. In the step lamp 61, the π boat selects β decoding in the encoding mode table of FIG. 24. " 1 is to decode the encoding bit stream 213 command selection information 200 Shima. In step ST62, the conversion unit 82 judges whether the coding mode table selection information 200 has been decoded according to the decoded coding mode table selection information 200 '------------ — (Please read the notes on the back before filling this page) 1T ^ ----- _ :: ”: LL_

”IJJ ~~: ______ 35 This paper ruler Central Standard of the Ministry of Economic Affairs 印 Printed by employee consumer cooperatives A7 B7. V. Description of Invention (33) Set to 0 and convert the output. Here, when the conversion unit 82 determines that the encoding mode table selection information 200 is set to 0, the encoded bit stream 213 is output to the encoding mode table A decoding unit 83. In step ST63, the encoding mode table A decoding unit 83 is Decoding encoding mode information 20 using the encoding mode table a. When the conversion unit 82 determines that the encoding mode table selection information 200 is not set to 1, the encoding bit stream 213 is output to the encoding mode table B and decoded. The unit 84 decodes the encoding mode table b grasshopper encoding mode information 206 in step ST64. In step ST65, the quantization step size decoding unit 85 inputs the encoded bit stream 213 and the decoded encoding mode information 206 to determine whether the encoding mode information 206 indicates a change in the quantization step size. When the quantization step size is instructed to change, the quantization step size decoding unit 85 decodes the quantization step size information 207 in step ST66. When the quantization step size change is not instructed, the 'quantization step size decoding unit 85 sets the quantization step size information 207 to 0 in step ST67. In step ST68, the quantization step size decoding unit 85 determines that the output quantization step size is the value obtained in step ST66 or step ST67. In step ST69, the change information decoding section 86 inputs the encoded bit stream 213 and the decoded encoding mode information 206 to determine whether the encoding mode information 206 indicates that there is change information. When the change information is instructed, the 'change information decoding unit 86 decodes the change information 208 in step ST70. In the last step ST71, the compressed video data decoding unit 87 decodes the compressed video data 205 of the input coded bit stream 213 and rotates it out. Fig. 28 is a flowchart showing a change compensation process of the change compensation unit 74 ', that is, step ST54 of Fig. 26. This paper is again applicable to China National Standard (CNS) A4 specification (210X297 mm) --------- Γ ------------- line " (Please read the Note: Please fill in this page again) Printed by the Consumer Standards Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Description of the invention (34) In steps ST81, ST83, and ST85, the conversion section 91 in Figure 25 determines whether to instruct the coding mode information It is one of prediction methods 1 to 4. In step ST81, when it is determined that the encoding mode information 206 is prediction method 1, the change information 208 is sent to the prediction method 1 implementation section 92. In step ST82, the prediction method 1 implementation section 92 inputs the change information 208 and the predicted image. The regional in-day position information 214 generates a prediction image 217 according to the prediction method 1. Similarly, when it is determined in step ST83 that the encoding mode information 206 is the prediction mode 2, in step ST84, the prediction mode 2 implementation unit 93 generates a prediction image 217 based on the prediction mode 2. In step ST85, when it is determined that the encoding mode information 206 is the prediction method 3, in step ST86, the prediction method 3 implementation unit 94 generates a prediction image 217 based on the prediction method 3. Then, in step ST85, when it is determined that the encoding mode information 206 is the prediction mode 4, in step ST87, the prediction mode 4 implementation unit 95 generates a prediction video 217 based on the prediction mode 4. As described above, if the image decoding device according to Embodiment 4 receives the bit stream obtained by multiplexing the encoding mode table selection information and the encoding mode information, and selects the encoding mode table decoding unit according to the encoding mode table selection information, The encoding mode information is converted into a prediction method and decoded. Even when the encoding bit stream has a small amount of encoding mode information in the additional information, it can be correctly decoded. Under the given conditions, it can still achieve good efficiency. Decoding device for decoding. 37 (Please read the notes on the back before filling out this page)

—W 定 线 #-This paper size is applicable to Chinese National Standard (CNS) A4 specification (210X 297 mm) A7 Printed by the Consumer Consumer Cooperative of the Central / Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the invention (35) Example 5 Example The image decoding device of 5 is the embodiment 4 of the MPEG-4 specification 5 which is improved to the 23rd image decoding device. The image coding corresponding to the __4 specification in Example 2: setting ::: should be coded device. The image solution of the clothes is shown in FIG. 29, which is the image decoding image of the fifth embodiment. In the figure, 101 series of syntax analysis • variable length coding 2, block code section, called inverse DCT section, _ series change compensation section: ^ The same reference number is the structure shown in Figure 23 of Embodiment 4 Fig. 23 shows the difference from the decoding device of the fourth embodiment. In the fifth embodiment, "the inverse DCT unit 10 is provided as the inverse orthogonal transform unit, and the sentence exaggeration analysis and the variable length encoding unit 1G1 From the exhausted bit stream 316, PE3 0.?, Which is the encoding, block data 307, which is the compressed image data, and compressed shape material 301, are decoded. Then, the shape decoding unit 102 unique to MPEG_4 is reset, and the shape decoding unit 102 decodes the compressed shape data 301 to obtain the decoded shape data 318. This decoded shape data 318 is configured by being reflected by the inverse DCT section 103 and the change compensation section 104. Even if 疋 MPEG-4 uses a rectangular video frame, such as a video signal as a coded video, the shape data must be rectangular, so the shape decoding unit 102 is not required at this time. Fig. 30 is a block diagram showing the structure of the syntax analysis and variable-length decoding unit 101 of the fifth embodiment shown in Fig. 29. In the figure, ιη series MBTYPE table selection information decoding section, 112 series shape data decoding section, 38 paper sizes are applicable to China National Standard (CNS) A4 specifications (210X297 mm) (Please read the precautions on the back before filling in this Page) .W. -Order. Printed by the Consumers Cooperative of the Central Standards Bureau of the Ministry of Online Economics A'7 ----------- -B7 Five 'Invention Description (36)' 113 MODB Decoding Department, 114 System conversion section, 115 series crossover data setting section, 116 series conversion section, 117 series code section, 118 series MBTYPE-2 decoding section, i 丨 9 series conversion section, i20 series cBpB decoding section, and 121 series CBPB return to zero Setting section, 122 series filling fast reading section,

123 series conversion section, 124 series DQUANT decoding section, 125 series DQUANT zero reset setting section, 126 series addition section, 127 series change information code section, 128 series addition section, 130 series block data decoding section, 131 series block data return Zero setting section. »The MBTYPE-1 table of the MBTYPE-1 decoding section 117 and the MBTYPE-2 table of the MBTYPE-2 decoding section 118 each have the contents of FIG. 9 and FIG. 10 described by the encoding device of the second embodiment. Fig. 31 is a block diagram showing the configuration of the change compensation unit 104 of the fifth embodiment shown in Fig. 29. In the figure, 14 丨 is a conversion section, 142 is a direct prediction section, 143 is a forward direction prediction section, 144 is a backward direction prediction section, and 145 is a two-direction prediction section. Next, clarify its actions. Fig. 32 is a flowchart showing the operation of the decoding device according to the fifth embodiment. The differences from the flowchart of the operation of the decoding device according to the fourth embodiment shown in Fig. 26 will be described. Since the fifth embodiment is composed of a decoding device corresponding to MPEG-4 having the shape decomposing unit 102 shown in FIG. 29, between steps ST51 and 8d52, step 8d51-1 is added to pass The shape encoding unit 102 performs shape decoding processing in units of microblocks, and processes in steps st53 and ST55 are inverse DCT processing, respectively. MBTYPE's judgment processing for whether to instruct mutual encoding is only a name change. Each step is essentially corresponding to 39 copies. The paper is again applicable to the Chinese National Standard (CNS) M specification (210X 297 mm) (Please read the notes on the back before filling out this page) Order A7 A7 Printed by the Central Consumers Bureau of the Ministry of Economic Affairs Consumer Cooperatives V. Invention Description (37 Steps ST53 and ST55 in FIG. 26 perform the same processing. The inverse DCT process in step ST53 and the change compensation process in step ST54 are performed in the shape decoding process in step ST51-1, and the decoded shape data 318 output from the shape decoding unit 102 is reflected, and each process is performed. As in the case of the fourth embodiment shown in FIG. 26, the inverse quantization processing and inverse DCT processing of steps ST52 and ST53, and the change compensation processing of step ST54 need not follow this order, even if they are reversed. The order, of course, can also be f lines at the same time. The parsing process of step ST51 usually detects the start code of each section of units such as VO (video object), VOL (video object layer), and VOP (video object plane), and performs decoding at this level. Here, The syntactic analysis processing of the system only deals with the processing of the microblock data, especially the stage analysis of the B-VOP data analysis. The MBTYPE table selection information 302 in FIG. 30 is used for decoding higher-level data such as VOP and VOL than microblocks. Fig. 33 is a flowchart showing the operation of the syntax analysis / variable length decoding unit 101, that is, the syntax analysis and variable length decoding processing of step ST51 shown in Fig. 32. First, in step ST91, the MBTYPE table selection information decoding unit 111 decodes the MBTYPE table selection information (B-VOP mode type) 302 of the B-VOP in the encoded bit stream 316 shown in FIG. In the next step ST92, the MODE decoding unit 113 decodes the multiplexed MODB (crossover decision information, not shown in Fig. 12) as additional information of the encoded bit stream 316. In the subsequent step ST93, the conversion unit 114 is used to determine whether the decoded MODB (crossover determination information) is 0 or not. 40 This paper size applies Chinese National Standard (CNS) A4 (210X297 mm) (Please read the precautions on the back before filling this page) Printed by the Consumers Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs " '1 _ '___— 丨 · V. Explanation of the invention (38) Here, when MODB is judged as 0 (NO in step st93), because this microblock is a span, in step ST94, the data setting unit at the time of the span The 115 series performs the following 1) to 3) as the setting processing at the time of span. 1) MBTQPEE = 0 (direct prediction). The MVDB (differential vector for direct prediction) is zero. 2) CBPB (showing all the information about the presence or absence of 8) and 8) of the encoding of the block as the DCT implementation unit in the microblock j) CT coefficients are all zero. In other words, the DCT coefficient data is all zero. 3) From the above two conditions, according to the condition that the MVDB is zeroed, the direct prediction 'predicted image is the original decoded image. In contrast, when the MODB is judged not to be 0 (YES in step ST93), because this microblock does not span, in step ST95, the conversion unit 116 judges the MBTYPE table selection information decoded in step ST91. Whether 302 is 0. When the MBTYPE table selection information 302 is 0 (YES in step ST95), in step ST96, the MBTYPE-1 decoding unit 117 selects the MBTYPE-1 table and performs MBTYPE decoding. In step ST97, the decoding sequence I shown below is performed. On the other hand, when the MBTYPE table selection information 302 is not 0 (NO in step ST95), in step ST98, the MBTYPE-2 decoding unit 118 selects the MBTYPE-2 table and performs MBTYPE decoding. In step ST99 ', the decoding order Π shown below is performed. In the final step ST100, the block data decoding unit 130 outputs the compressed image 41. The paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) ------------ (please first Please read the notes on the back to fill in this paper.) "? Τ Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs A7 _____ B7 V. Invention Description (39) ~ 'Block data 307 of the data. Fig. 34 is a flowchart showing the decoding sequence I of step ST97 shown in Fig. 33. According to this process, first in step ST101, the conversion unit 119 of FIG. 30 detects the MBTYPE determination (1) process, that is, it is set as the cipher of the encoding mode information (MBTYPE) 308 in the encoded bit stream. Referring to the MBTYPE-1 table shown in FIG. 9, it is determined whether the mode number (the value of MBTYPE) corresponding to the detected cipher phrase is 3. Here, when the pattern match code is 3, refer to the μβτυρε ^ table shown in FIG. 9 because in the case of a filled coding mode, step STi0i is,, 疋, in the subsequent step ST102, fill The fast reading unit 122 performs fast reading processing by using the compressed image data of the encoding mode as a filling bit, and returns to the decoding processing of MODB (crossing judgment information) in step ST92 in FIG. 33. On the one hand, when the coding mode is not 3, refer to the MBTYPE-1 table shown in FIG. 9 because it is a coding mode that is not a padding mode. In this step, stioi is “No”, and in the subsequent step ST103. The intermediate conversion department b judges whether the MODB (crossover judgment information) is 2. Here, 'when MODB is 1,' in step ST104, the CBPB zero resetting unit 121 will display all the DCT coefficients of the encoding of the child block of the 8 X 8 block of the DCT implementation unit in the microblock. The CBPB of all information is set to zero. On the other hand, when MODB is 2, in step ST105, the CBPB decoding unit I20 decodes the CBPB from the encoded bit stream 316. In the subsequent step ST106, the conversion unit 123 refers to the MBTγρΕ 42 paper standard applicable to the Chinese national standard 8 4 Lin (--- {Please read the precautions on the back before filling this page} _w · 线 线 #-· A7 B7 Printed by the Employees' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the Invention (40) Decision (1) Processing, which is the μβτυρε ^ table shown in Figure 9, determines that the encoding mode corresponds to the encoding mode set in the encoding bit stream 316 Is the mode number of the secret word of the information (MBTYPE308) 1 or 2. When MBTYPE is 1, 2 'In step ST107, the DQUANT decoding unit 124 decodes DQUANT (quantization step difference value). From the bit stream. When MBTYPE is not 1, 2, the DqUANT zero setting unit sets DQUANT to zero in step 108. In step ST10g, the 'addition unit 126 resets the DQUANT decoding unit 124 or the DQUANT zero setting unit 125. The output is added to the quantized step size information of the nearby microblock to output the quantized step size information 309. In step stiio, the change information decoding unit 127 refers to the MBTYPE-1 determination and processing, which is the MBTYPE-1 table, Judgment corresponds to being set to coded bits Is the mode number of the cipher word of the encoding mode information (MBTYPE308) in the stream 316 2. When MBTYPE is 2, 'in step ST111, the change information decoding unit 127 converts the MVDf (forward direction prediction change vector difference value) from the bit stream. ) Decoding, the addition unit 128 adds the MVDf and the change information of the nearby microblocks, and outputs the change information 310 for forward prediction. In step ST112, the change information decoding unit 27 determines with reference to MBTYPE-1 (4 ) Processing ', that is, the MBTYPE-1 table, determines whether the mode number corresponding to the cipher word set as the encoding mode information (MBTYPE308) in the encoded bit stream 316 is 1. In step ST113, the change information decoding unit 127 follows from Decode the MVDb (change vector difference value for backward direction prediction) in the bit stream, the adder 128 and the nearby MVDb (please read the precautions on the back before filling this page), w 定 线 # This paper size is applicable to China Standard (CNS) A4 specification (210X297 mm) Printed by the Consumer Standards Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs A7 B7 V. Invention information (41) The micro-block change information is added together, and the output is the change for the backward prediction Information 310. In step ST114, the change information decoding unit 127 refers to the MBTYPE-1 determination (5) process, that is, the MBTYPE-1 table, and determines that it corresponds to the encoding mode information (MBTYPE308 set in the encoding bit stream 316). ) Whether the mode number of the cipher is 0. When MBTYPE is 0, in step ST115, the change information decoding unit 127 decodes the MVDB (direct prediction, using a difference vector) from the bit stream, and restores change information 310 for direct prediction, such as a direct prediction vector. The decoding process shown in Fig. 34 above ends. Step ST109 is after steps ST107 and ST108. Even after step ST115, the processing of the addition unit 128 of steps ST111 and ST113 may be performed simultaneously with or before step ST109, or may be performed simultaneously with step ST100 of FIG. 33. Or later. If the processing of step ST115 is completed, in the last step ST100 of FIG. 33, the conversion unit 129 judges the CBPB (displayed as the DCT implementation unit in the microblock) output from the CBPB decoding unit 120 or the CBPB zero setting unit 121. 8 X 8 block of the sub-block encoding all the information about the existence of DCT coefficients). When the CBPB is a zero block, the block data zero setting section 131 outputs all the block data (DCT coefficient data) to zero. On the other hand, when the CBPB is not a zero block, the block data decoding unit 130 outputs the block data decoded with variable length. Fig. 35 is a flowchart showing the decoding sequence ** .., II of step ST99 shown in Fig. 33. 44. This paper size applies Chinese National Standard (CNS) A4 specification (210X 297 mm) (Please read the notes on the back before filling this page) Order A7 B7 Printed by the Consumers Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs (42) First, in step ST121, the conversion unit 120 detects the MBTγρΕ_2 decision (1) processing, that is, it is set as the encoding mode in the encoding bit stream 316-the secret of MBTYPE 3 0 8 'reference The MBTγρE2 table shown in FIG. 10 determines whether the pattern number corresponding to the detected secret word is 8. When the mode number is 8, refer to the MBTγρΕ_2 table shown in FIG. 10, because in the case of a filled coding mode, step ST121 is, Yes, in the subsequent step ST122, the fast reading unit 122 is filled. The compressed scene and image data in this coding mode are used as filling bits, and a fast reading process is performed, and the process returns to the MODB (cross decision information) decoding process in step ST92 in FIG. 33. On the one hand, when the encoding mode is not 8, refer to the MBTYPE-2 table shown in FIG. 10 because it is a case where the encoding mode is not the padding mode. This step ST121 is “No”, and in the subsequent step ST123 The conversion unit 119 determines whether the MODB (crossover determination information) is 2. Here, when MODB is 1, as in the case shown in FIG. 34, in step ST124, the CBPB zero setting unit 121 is a sub-block of 8 × 8 blocks that displays the DCT implementation unit in the microblock. The CBPB of all the information about the presence or absence of the DCT coefficients of the encoding is set to zero. On the other hand, when MODB is 2, 'in step ST125, the CBPB decoding unit 120 exhausts the CBPB from the encoded bit stream 316. In the following step ST126, the conversion unit 123 refers to the MBTYPE-2 decision (2) processing, that is, the MBTγρE_2 table shown in FIG. 10, and determines that the encoding mode corresponds to the encoding mode set in the encoding bit stream 3 丨 6. Is the mode number of the secret word of the information (ΜΒΤΥΡΕ308) i, 3, 5, 7? This paper size is applicable to Chinese National Standard (CNS) A4 specification (2T0X297). (Please read the notes on the back and fill in this form. / Λ. Printed by the Consumer Standards Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 5 (43) When the MBTYPE is 1, 3, 5, or 7, in step ST127, the DQUANT decoding unit 124 decodes DQUANT (quantized step size difference value) from the bit stream. It is not 1 in MBTYPE308. At 3, 5, 5, and 7, in step 128, the DQUANT zeroing setting unit 125 sets DQUANT to zero.-In step ST129, the addition unit 126 sets the output of the DQUANT decoding unit 124 or the DQUANT zeroing setting unit 125 to the vicinity. The quantized step size information of the microblocks are added to output quantized step size information 309. In step ST130, the change information decoding unit 127 refers to the MBTYPE-2 decision (3) processing, that is, the MBTYPE-2 table, Determine whether the mode number corresponding to the cipher phrase set as the encoding mode information (MBTYPE308) in the encoded bit stream 316 is 2, 3, 6, or 7. When MBTYPE is 2, 3, 6, or 7, in step ST131 The change information decoding unit 127 converts the MVDf (forward direction prediction) from the bit stream The change vector difference value) is decoded, and the addition unit 128 adds the MVDf and the change information of the nearby microblocks and outputs the change information as forward prediction change information 31. In step STI32, the 'change information decoding unit 2' is a reference. The MBTYPE-2 determination process, that is, the MBTYPE-2 table, determines whether the mode number corresponding to the cipher phrase of the encoding mode information (MBTYPE308) in the encoding bit stream 3 16 is 2, 3, 4, or 5. In step ST133, the change information decoding unit 27 decodes the MVDb (change vector difference value for backward direction prediction) from the bit stream, and the addition unit 128 adds the MVDb and change information of the nearby microblock to output It is the change information 310 for backward direction prediction. The decoding process shown in the above figure is as shown in Figure 46. 46 This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) ------ ^ --- 1C- ------ IT ------ # 11.- B (Please read the notes on the back before filling out this page) Printed by the Consumers' Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs 5. Description of the Invention (44 bundles). If the processing of step STl33 is performed, the conversion unit 129 judges: the beam is in the last step of 帛 33®

It is judged that the C coffee that is unloaded from the CBPB decoding unit 120 or cBPB to β and 疋 ° 8 rounds (shown as the DCT implementation unit in the microblock = X "All the coefficients of the sub-blocks of the block are positive. Information). When the CBPft base office 1Q1 _ _ is zero, the block data is reset to zero. The unit 131 outputs block data γ a > ^ juice (DCT coefficient data) 307 to all zeros. On the one hand, the block data decoding unit 130 outputs block data 307 that is variable-length decoded when the CBPB stone rabbit Christine, and the horse are zero blocks. Fig. 36 is a flowchart showing the operation of the non-change compensation unit, that is, the change compensation process of step ST54 shown in Fig. N. The change compensation process of the fifth embodiment shown in the first embodiment is the embodiment of the prediction methods 1 to 4 of the change compensation process of the first embodiment * shown in the first embodiment. Change makes up. The conversion unit 114 of P 104, in steps ST141 and ST143 ST145, determines whether the coding mode is instructed to be any of direct prediction, forward prediction, backward prediction, and both directions based on the encoding mode information (MBτγρΕ ·). In step ST141, when it is determined that the conversion unit 141 indicates that the encoding mode is direct prediction, the change information 31o is sent to the direct prediction unit M2. In step ST142, the direct prediction unit 142 is based on the input change information 31o and Direct prediction of the location information 317 in the daytime plane of the predicted image area generates a predicted image 322. In step ST143, when the judgment conversion unit ι41 indicates that the encoding mode is forward prediction, similarly in step ST144, the forward direction prediction unit 143 applies the Chinese national standard (CNS) A4 specification (21〇χ 297 mm) to the paper size. ) (Please read the precautions on the back before filling this page) Threading ®-Printed by A7 B7, Employee Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 5. Description of the invention (45) The prediction image 322 is generated according to the forward direction prediction. In step ST145, when the judgment conversion unit 141 instructs the encoding mode to be backward prediction, in step ST146, the backward orientation prediction unit 144 generates a prediction image 322 based on the backward prediction. When it is judged that the encoding mode is not instructed to backward prediction In step ST147, the two-direction prediction unit 145 generates a prediction image 322 based on the two-direction prediction. As described above, according to Embodiment 5, in the encoding device side, from the plural encoding mode tables of the MBTYPE-1 table and the MBTYPE-2 table that define a combination of appropriate encoding modes that correspond to the target bit rate, etc., Select the most suitable encoding mode table. Even when data is transmitted, MBTYPE-1 decoding unit 117 and MBTYPE-2 decoding unit 118 are set to decode each corresponding to the plurality of encoding mode tables, according to the encoding bit on the encoding device side. MBTYPE table selection information 302 and encoding mode information (MBTYPE308) of encoding mode selection information in the meta stream 316. MBTYPE-1 decoding unit 117 and MBTYPE-2 decoding unit 118 are selected. Since the decoding is performed according to the encoding mode during encoding, the encoding is performed according to the encoding. The complex coding mode table on the device can correctly decode even if the amount of coding mode information in the additional information of the coded data is small. Under the given conditions, a decoding device capable of efficient decoding can be realized. Embodiment 6 Fig. 37 is a block diagram showing the configuration of an image decoding device according to Embodiment 6. The image decoding device of the sixth embodiment is described in more detail in the configuration of the inverse quantization unit of the image decoding device of the fifth embodiment corresponding to the MPEG-4 specification shown in FIG. 29, which corresponds to the structure of FIG. Example 3 48 This paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm) (Please read the notes on the back before filling out this page) Threading ·-A7 Printed by the Staff Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs- ---_ V. An image decoding device of the image coding device of the invention description (46). In the figure, 151 is a parsing and variable-length decoding unit, 152 is a conversion unit that converts the selected bit type 402 according to the existing bitch method decoded from the encoded bit stream 416, and 153 is a low bit rate. The H.263 type inverse quantization unit corresponding to the H.263 standard, 154 is an MPEG_2 type inverse quantization corresponding to the MPEG-2 standard using a high bit rate bit rate. Since the other configuration is the same as that of the fifth embodiment shown in FIG. 29, the same reference numerals are used to designate the same reference numerals, and descriptions thereof are omitted. Fig. 38 is a block diagram showing the structure of the syntax analysis and variable length decoding unit 151 of the sixth embodiment shown in Fig. 37. The syntax analysis and variable-length decoding section 151 is basically the same as the syntax analysis and variable-length decoding section 101 of the fifth embodiment shown in FIG. 30. The difference is only as shown in the figure. Decoding section 16] [Decodes the existing video object layer quant type 4 2 in the encoding bit oil 416 shown in FIG. 22, and the conversion section 162 selects the information 402 instead of the quantization method 402 MBTYPE selection information, convert MBTYPE] decoding unit 117 and MBTYPE-2 decoding unit 118. The decoded quantization method selection information 402 is also output from the syntax analysis and variable length decoding unit 151 to the conversion unit 152 in the π-th graph. The other structures are the same as those of the syntax analysis / variable length decoding unit 101 of the fifth embodiment shown in FIG. 30, and therefore are denoted by the same reference numerals, and descriptions thereof are omitted. Fig. 39 is a flowchart showing the operation of the image dissolving device according to the sixth embodiment. The operation of Example 6 is basically the same as that of Example 5 shown in Figure 32 (please read the precautions on the back before filling this page). Alignment »-1 This paper size applies the Chinese National Standard (CNS) A4 specification ( 2 丨 OX25 »7mm) A7 B7 printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs 5. The description of the invention (47) is the same, and the difference is explained in step 151. Syntax analysis and encoding data decoding processing, and selection Processes other than steps ST152 to ST154 of the inverse quantization of the quantization method. In summary, according to Embodiment 6, since the quantization method is converted, in step ST152, the conversion unit 152 of FIG. 37 determines the quantization based on the existing quantization method selection information 402 decoded from the encoded bit stream 416, and judges the quantization. Whether the method is H.263. When the quantization method is H.263, the process proceeds to step ST153, and the H.263 type inverse quantization unit 153 advances the 4th inverse 1: subization according to the MBTYPE-1 table corresponding to the low bit rate. On the other hand, when the quantization method selection information 402 is MPEG-2, the process proceeds to step ST154, and the MPEG-2 type inverse quantization unit 154 performs inverse quantization according to the MBTYPE-2 table corresponding to the high bit rate. Fig. 40 shows the operation of the syntax analysis and variable-length decoding unit 151, that is, the flowchart of the syntax analysis and encoding data decoding process of step ST151 shown in Fig. 39 '---------- The processing shown in FIG. 40 is basically the same operation as the processing of the fifth embodiment shown in FIG. 33 except that the quantization method selection information 402 is used as the MBTYPE table selection information. First, in the first step ST161, the quantization method selection information decoding unit 161 decodes the quantization method selection information 402 from the encoded bit stream 416. In step ST162, the conversion unit 162 selects the information 402 according to the quantization method, and determines whether the quantization during encoding is H.263 type quantization. If it is H.263 type quantization, in step ST96, the paper size is 50 Applicable to China National Standard (CNS) Α4 specification (210 × 297 mm) (Please read the precautions on the back before filling this page)

1. Line 1T Printed by the Consumer Cooperatives of the Central Bureau of Standards, Ministry of Economic Affairs A7 ____B7____ 5. Description of the Invention (48) The MBTYPE-1 decoding unit 117 selects the MBTYPE-1 encoding mode table for decoding. On the other hand, if it is not quantization of the H.263 type, but quantization of the MPEG-2 type, in step ST98, the MBTYPE-2 decoding unit 118 selects the MBTYPE-2 encoding mode table for decoding. As described above, 'If according to Embodiment 6, the parts other than the conversion of the quantization method are the same as the image decoding device of Embodiment 5 above, and the same as the case of Embodiment 5 above, according to the plurality of encoding device side; code. ... one, * ---- mode table, even if the amount of encoding mode information in the additional information of the exhausted code data is small, it can be correctly decoded, because the existing VOL header information is used The MBTYPE table selection information (B VOP mode type) of the quantization method selection information 402 as the encoding mode group selection information. There is no need to add any additional information in the current syntax of the encoded bit stream 416. Under the given conditions, A decoding device capable of performing efficient decoding is realized. In the sixth embodiment, the MBTYPE-1 table and the MBTYPE-2 table are taken as an example. As described in the second embodiment, as the other, as the MBYTYPE table corresponding to the low bit rate, for example, MBTPPE-0, MBTYPE-3 are used. , MBTYPE-5, MBTYPE-6, MBTYPE-7, etc. In addition, as the MBYTYPE table corresponding to the high bit rate, for example, it is also possible to use an encoding mode table such as MBTYPE-4. The industrial applicability is as described above. In some of the image encoding method, image encoding device, image decoding method, and image decoding device of the present invention, even if the encoding mode is increased, the amount of encoding mode information in the additional information of the electronically encoded data will not increase. , Suitable for efficient transmission. 51 * (Please read the notes on the back before filling out this page) Threading ·-This paper size applies to China National Standard (CNS) A4 (210X297 mm) ~

Claims (1)

A8 B8 C8 D8 VI. Patent application scope 1. An image coding method that converts the coding mode on a specified area and simultaneously encodes the input image. According to the multiple coding mode groups that may select the dragon coding mode according to the registration in each area Select the coding mode and information group specified in the information, select the coding mode group, and according to the selected coding mode 1, 1, ... -Group, which encodes the above-mentioned input image on each region to generate a compiled bit stream, and appends the coding mode group selection information to the encoded bit stream and outputs it. 2 · —An image encoding device, including an image encoding device that converts an encoding mode on a predetermined area and encodes an input image at the same time, according to ----._________________ · <: ........ ... record the complex coding mode group and the predetermined coding mode group selection information that may select the coding mode, select the coding mode group editor for coding processing ___ one by one ... Touch the group and select the device. Based on the selected coding mode group, from the registered coding modes (........, '., ...), select the coding of the coding mode in the above areas. The mode selection device, according to the selected thin weft code mode, encodes the above-mentioned area and outputs encoding data, and encodes the above into the Lido group. Selection information '......... --------------------------------------------------- ------ " ....... The multiplexing device that multiplexes the above-mentioned encoding mode and the above-mentioned encoding data, and outputs a coded bit stream. 3. As described in item 2 of the scope of patent application Image coding device, _ Among them, the consumer cooperation of the Central Bureau of Standards of the Ministry of Economic Affairs Printed by the company (please read the precautions on the back before filling out this page). The number of coding mode groups will register the coding mode that may be selected in the coded image list. ______. _______; — Fill in the micro area ▲, coding mode group selection The device is based on the video object plane composed of the above-mentioned microblocks of the image representing the state of the video object at each moment, and selects the coding mode to be used for coding from the complex editing mode group based on the selection information of the cotton code mode group. Group. '4. For example, the image scale device of item 2 of the scope of patent application, of which 52 paper sizes are applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) ... 2 sets include different sets of quantization I sets, and Gen Behr selects the above-mentioned complex quantization methods. 1 When selecting the set, the above selection is used when encoding, and the encoding mode group selection device uses == for quantization, Select information for _ pattern group. The sub method selects information to be the low application level of the coding mode corresponding to the low bit rate of the t4 car of the T4 car of the chest of the material shirt. And the register bit pattern corresponding to the bit pattern that is higher than the predetermined reference bit is composed of a coding pattern group. , ..., encoding mode image decoding method: 'Qian Ren squeezes the image' into the stream, and the shadow is edited in the specified area] 蜗 'from the above-mentioned editing Xianyuan She will edit each region of the Wei-style Gunma method Shout recording possible _ _ _ 'In the support mode group, select the above-mentioned editing type group to select your newsletter ^ transcode the selected encoding mode group' to decode from the encoded data of each area of the editing mode. ① In-machine pair 7. A kind of image decoding device, including inputting a bit field for encoding and encoding the image in a predetermined area: "Compression of a compression codec" When encoding the above areas, Deng Feng tile can select the image Decoding Pei group, selecting information decoding equipment from the encoding mode group for information decoding in the above coding bit complement, and selecting the encoding mode group indicated by the above encoding mode group selection information from the complex encoding group: The selected coding mode group, --------- 1C ------, 玎 ------- line Φ- (Please read the precautions on the back before filling. ¾ This page} A8 B8 C8 D8 VI. Patent application scope In the above-mentioned encoding bit stream, an encoding mode decoding device that decodes the encoding mode used for encoding the above-mentioned regions, and according to the encoding mode, each region is decoded from the above-mentioned encoding bit-stream. Decoding device for decoding encoded data. 8. For example, the image decoding device under the scope of patent application No. 7, in which the complex encoding mode group may be repeatedly encoded-encoded to it in the encoded image list •.-.. ..... On micro-blocks, encoding mode The avoidance selection device is based on the image of the worm representing the video object at each moment, and the video object edited by the remote and micro-area pain is selected on the surface. Based on the coding mode group, it selects information from the complex coding mode group (in the above). Select the encoding mode group to be used for decoding. 9. For example, the image decoding device in the seventh scope of the patent application, wherein the decoding device includes a plurality of different inverse quantization devices, and the complex inverse quantization method is selected according to the quantization method [-selection information. Any of the two inverse quantization selection devices, when encoding, use the selected inverse-quantization device to perform inverse quantization, and the encoding mode group selection device uses the cliff quantization method to select information as Coding mode group selection information. 10. For example, the image decoding device in item 7 to 9 of item 2 of the patent application range, wherein the complex coding mode group m records a bit rate corresponding to a bit rate lower than a predetermined reference bit rate. The low bit rate corresponding to the coding mode is printed. Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs (please read the precautions on the back before filling in (Page) t group, and registration mode group of high bit rate corresponding coding mode registered with bit rate higher than the specified reference bit rate. This paper size applies Chinese National Standard (CNS) A4 specification (210X297 mm) )