JP2010045857A - Image encoding method, device, and image processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image encoding method and apparatus, encoding a motion image enabling random access thereto, with low delay and at high compression. <P>SOLUTION: An image encoding method encodes image data using a reference image about the input image data and a difference between the reference image and the image data. The method executes a procedure for setting an image output possibility parameter which shows a possibility of decoding and outputting of the image data so long as the images in the past predepredetemindtermined range have been decoded when the image data are decoded, a procedure for calculating effect range information showing the range of images, encoded in the past to be effected when the image data are encoded, and a procedure for controlling the range of reference images at encoding so that calculated effect range information does not exceed the image output possibility parameter. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a technique for encoding / decoding a moving picture, and particularly to a technique effective when applied to a method and apparatus for encoding / decoding a moving picture with a small delay time.

  Conventionally known moving image encoding techniques represented by JPEG and MPEG-2 use the characteristic that image information generally has, that is, the characteristic that there is a high correlation between adjacent pixels and frames, Furthermore, by reducing redundant information such as high-frequency components that are difficult for human vision to perceive changes, an input image in a wide transmission band (for example, about 1.5 Gbps in HD-SDI (High Definition Serial Digital Interface)) can be obtained. It can be applied to a low transmission band (for example, about 15 Mbps in terrestrial digital broadcasting).

  As such a technique, for example, JP-A-5-300492 (Patent Document 1), JP-A-11-4437 (Patent Document 2), JP-A-11-234671 (Patent Document 3), JP Techniques such as 2003-348538 (Patent Document 4) have been proposed.

  In recent years, H.264 has been established by a joint project of ISO / MPEG (International Organization for Standardization / Moving Picture Experts Group) and ITU-T / VCEG (International Telecommunication Union-Telecommunication standardization sector / Video Coding Experts Group). The H.264 / AVC image coding system has been widely used due to its compression rate performance. This H. As for the H.264 / AVC standard, for example, Joint Video Team (JVT) of ISO / IEC MPEG & ITU-T VCEG: Text of ISO / IEC 14496-10 Advanced Video Coding 3rd Edition (2004). Are described in detail.

JP-A-5-300492 Japanese Patent Laid-Open No. 11-4437 JP-A-11-234671 JP 2003-348538 A

Joint Video Team (JVT) of ISO / IEC MPEG & ITU-T VCEG: Text of ISO / IEC 14496-10 Advanced Video Coding 3rd Edition (2004).

  By the way, the above H.P. In the H.264 / AVC standard, encoding processing is performed in units of 16 × 16 pixel blocks called macroblocks. In intra-screen or inter-screen prediction, it is specified that a macro block is divided into blocks of 16 × 16 pixels, 8 × 8 pixels, 4 × 4 pixels, etc., and encoding is performed by prediction processing for each block unit. ing. According to this technique, encoding efficiency can be improved by properly using a large number of defined prediction modes in accordance with fine movements or patterns of an input image.

  Hereinafter, H.C. An outline of the intra-screen prediction process and the inter-screen prediction process will be described using the H.264 / AVC standard as an example.

  In-screen prediction is a method of improving the compression rate by using pixel correlation between adjacent pixels in the same screen. For example, Intra4 × 4DC prediction processing, which is one of the intra prediction modes, is prediction in units of 4 × 4 pixels, and includes 4 pixels adjacent to the upper side of the corresponding 4 × 4 pixel block and 4 pixels adjacent to the left side. An average value of a total of 8 pixels is used as a predicted value. The difference between each pixel value of the corresponding block and the predicted value is converted, subjected to quantization and variable length coding, and stored in the bitstream. H. The intra prediction mode in H.264 / AVC is set to 8 modes in units of 4 × 4 pixels, 8 modes in units of 8 × 8 pixels, and 4 modes in units of 16 × 16 pixels. Note that the above is related to luminance data, and four modes are set separately from luminance for color difference data.

  Inter-screen prediction is a method of improving the compression rate by using pixel correlation between screens of images. For example, in the P4 × 4 prediction process, which is one of the forward prediction modes, a prediction image is generated using a similar region of a picture that is positioned forward (past) in the processing time with respect to an encoding target block. Inter-screen prediction has a plurality of prediction modes corresponding to the prediction pixel size, and a prediction image is generated according to the modes. H. In H.264 / AVC, 8 types of forward prediction modes are available: 4 × 4 pixels, 8 × 4 pixels, 4 × 8 pixels, 8 × 8 pixels, 16 × 8 pixels, 8 × 16 pixels, 16 × 16 pixels, and skip. There is. In bidirectional prediction, there are 23 types of prediction modes. One of the modes with the smallest encoding cost value is selected from these, and encoding is performed.

  In general, with intra-screen prediction and inter-screen prediction, prediction values with higher accuracy are obtained with inter-screen prediction. Therefore, by increasing the inter-screen prediction usage rate, the difference between each pixel value and the predicted value can be reduced, and a high compression rate can be obtained. On the other hand, inter-screen prediction uses another screen to calculate a predicted value, and therefore cannot be normally decoded if the screen used for prediction at the time of decoding the target screen is not decoded. In general, a picture (I picture) configured only by intra prediction is periodically inserted to limit the range of pictures used for inter prediction at the time of decoding. As a result, random access is possible.

  For a video stream that does not require random access, only the first picture is set as an I picture, and the other pictures are configured as pictures including inter-screen prediction (P pictures). There are also ways to improve it. However, in this method, since the reproduction from the middle position cannot be performed, the usage method is limited.

  On the other hand, in image encoding, it is generally known that a delay occurs during encoding and decoding. After inputting an image to the image encoding device, the code output from the image encoding device is input to the image decoding device, and the decoding process is executed to display the same picture. Is a delay time. In a general image encoding and decoding apparatus, the delay is several hundred milliseconds to several seconds. Recently, for the purpose of simultaneous use with uncompressed image data and the like, there is a need to reduce the delay of image encoding and decoding such as a delay time of 16 milliseconds or less.

  In order to realize low-delay image encoding and decoding, it is necessary to suppress fluctuations in the amount of data generated. In general, image data is transmitted at a fixed bit rate determined from the data amount of the entire image data, and the transmitted image data is stored in a buffer and processed.

  Here, when a large amount of bits exceeding the bit rate is generated in a certain area of an image picture, the data in the buffer is not sufficient for the processing, and waiting for transmission of data to the buffer occurs. End up. If all the processes for the area are not completed, the area cannot be displayed. As a result, the delay time for displaying the entire image increases. That is, the delay time is set by the time for processing the assumed maximum bit generation amount, and this time is greatly influenced by the difference between the bit rate and the maximum bit generation amount, that is, the amount of fluctuation of the generated data.

  In the conventional technique, it is conceivable to encode an image by the following three methods. First, in a general image encoding method, I pictures are periodically inserted between consecutive P pictures. Since the I picture has a larger amount of generated bits than the P picture, it is necessary to set the delay time in accordance with the processing of the I picture. In this case, the delay time is increased because the amount of generated data is large. For example, in terrestrial digital broadcasting, I pictures (I frames) are inserted at a rate of once every 0.5 seconds. Therefore, an encoding and decoding delay of about 500 milliseconds to several seconds has occurred.

  As another method, it is also possible to configure the code with all I pictures. In this case, since the bit rate is increased and the fluctuation of the data generation amount is small, the delay time can be reduced. However, since the I picture generally has a lower prediction accuracy than the P picture, the compression rate is low. When all of the I pictures are composed of I pictures, it is difficult to improve the compression rate.

  As another method, a configuration in which only the first picture is an I picture and the others are P pictures can be considered. However, as described above, playback from an intermediate position cannot be performed, and the usage method is very limited. End up.

  SUMMARY OF THE INVENTION An object of the present invention is to provide an image encoding method and apparatus capable of encoding a randomly accessible moving image with low delay and high compression.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in the present application, the outline of typical ones will be briefly described as follows.

  In the image coding method according to the representative embodiment of the present invention, a value that defines the maximum value of the range of a picture that is affected when an image is coded is set. It is characterized in that it is managed whether or not it is influenced by a past picture that is far away, and control is performed so that the range of the picture that is actually affected is less than the value set above.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  According to a typical embodiment of the present invention, it is possible to provide a moving image encoding method and apparatus that can be randomly accessed with low delay and high compression without requiring periodic I-picture insertion. It becomes.

It is the figure which showed the structure of the processing module in the image coding apparatus of Embodiment 1 of this invention. It is a flowchart showing the flow of a process of the encoding process module in the image coding apparatus of Embodiment 1 of this invention. It is a figure showing the outline | summary of the influence picture number information table in the image coding apparatus of Embodiment 1 of this invention. It is the conceptual diagram which arranged the picture in the order processed at the time of the image coding in the image coding apparatus of Embodiment 1 of this invention. It is a flowchart at the time of the setting of the image output start picture number Ns in the image coding apparatus of Embodiment 1 of this invention. It is a flowchart which shows the initialization operation | movement of the influence picture number information table in the image coding apparatus of Embodiment 1 of this invention. It is a flowchart which shows the operation | movement of the influence picture number control part at the time of the prediction process execution in the image coding apparatus of Embodiment 1 of this invention. It is a flowchart which shows operation | movement of the influence picture number control part at the time of the update of the influence picture number information table in the image coding apparatus of Embodiment 1 of this invention. It is the figure showing the outline | summary of the influence picture number information table in the image coding apparatus of Embodiment 2 of this invention. It is a conceptual diagram of the picture used for the reference in the image coding apparatus of Embodiment 2 of this invention. It is a flowchart which shows operation | movement of the influence picture number control part at the time of the prediction process execution in the image coding apparatus of Embodiment 2 of this invention. It is a flowchart which shows operation | movement of the influence picture number control part at the time of the update of the influence picture number information table in the image coding apparatus of Embodiment 2 of this invention.

<Outline of Embodiment of the Present Invention>
In the coding of moving images, the problem of coding with a P picture is that it is unclear because how much the processing of the corresponding picture is affected by each picture that is separated in time differs for each pixel or block, This is also a factor that prevents random access during decoding. If the presence / absence of the influence of a picture that is separated in time is clarified, the picture can be decoded without any problem by starting decoding from the affected picture, and random access becomes possible.

  Therefore, the image coding apparatus according to an embodiment of the present invention grasps the temporal distance between pictures based on the number of pictures calculated based on the difference in picture numbers, and the range of pictures that are affected when the image is coded. The image output start picture number Ns is set as an image output enabling parameter that can process the maximum number of pictures, that is, if this number of pictures is processed at the maximum, the corresponding picture can be decoded and output. Further, the influence picture number Ne is calculated and managed as influence range information indicating how much the corresponding picture is actually affected by past pictures during the prediction process, and the influence picture number Ne is output as an image. Control is performed so as to be smaller than the starting picture number Ns.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

<Embodiment 1>
Hereinafter, an image encoding apparatus (encoder) according to the first embodiment of the present invention will be described. FIG. 1 is a diagram showing a configuration of a processing module in the image coding apparatus according to the present embodiment. The image encoding apparatus is mounted on, for example, an LSI, and can be mounted on various image processing devices that perform moving image encoding processing, such as a television / video camera, a DVD / HDD recorder, a mobile phone, and a digital camera. . It is also possible to implement the function of each processing module as software that runs on a computer system.

  The encoding processing module in the image encoding apparatus according to the present embodiment includes a prediction unit 102, a transform / quantization unit 103, a variable length encoding unit 104, an inverse quantization / inverse transform unit 105, a frame memory 106, and the number of affected pictures. The configuration includes a control unit 108, an affected picture number information table 109, and an Ns storage area 110.

  The input image 101 is input to the prediction unit 102, and a prediction process is performed. As a result of the prediction process, a prediction image is output from the prediction unit 102, and a difference from the input image 101 is output to the transform / quantization unit 103. The data converted and quantized by the conversion / quantization unit 103 is encoded by the variable length encoding unit 104 and output as a bit stream 107. On the other hand, the transformed / quantized data is inverse-quantized / inverse-transformed by the inverse-quantization / inverse-conversion unit 105 and then added to the predicted image, and is stored in the frame memory 106 for prediction processing in the next stage. Stored.

  The Ns storage area 110 is a means for holding the image output start picture number Ns111 that is an image output enable parameter, and the affected picture number information table 109 is a means for holding the affected picture number Ne that is the affected range information. The picture number control unit 108 is a means for controlling the range of the reference picture during the prediction process in encoding.

  When the prediction process is executed, the prediction unit 102 obtains information on whether or not the reference picture area used for the prediction process is usable from the affected picture number control unit 108, and when it is unusable (NG), do not use. The affected picture number control unit 108 acquires area information of the reference picture used for the prediction process from the prediction unit 102, and acquires affected picture number information of the corresponding area from the affected picture number information table 109. The acquired affected picture number information is compared with the image output start picture number Ns111 stored in the Ns storage area 110. If the affected picture number information is smaller, an OK signal is output to the predictor 102. To do.

  Next, the flow of processing in the image coding apparatus according to the present embodiment will be described. FIG. 2 is a flowchart showing the processing flow of the encoding processing module in the image encoding apparatus according to the present embodiment.

  First, the image encoding device is initialized (step 201). Next, the image output start picture number Ns111 is input to the Ns storage area 110 (step 202). Thereafter, the affected picture number information table 109 is initialized (step 203). In this case, the initial value is larger than the image output start picture number Ns111.

  The following part is looped until the encoding process is completed. First, the input image 101 is input (step 204), and the prediction unit 102 sets a prediction region (step 205). The prediction unit 102 obtains information on whether the prediction region is usable (OK) or unusable (NG) from the affected picture number control unit 108 (step 206), and performs the determination (step 207). In the case of OK, the prediction error of the prediction region is evaluated (step 208). In the case of NG, the prediction region is not evaluated, and then it is determined whether the evaluation of all prediction regions is completed. (Step 209). If the evaluation of all prediction areas is not completed, the process returns to the next prediction area setting (step 205). If the entire prediction area is unusable (NG), processing such as intra prediction is performed.

  After the evaluation of all the prediction areas, the influence picture number information table 109 is updated based on the determined prediction area information (step 210). Thereafter, a difference value between the input image 101 and the predicted value is generated (step 211), and transformation / quantization processing is performed on the difference value in the transformation / quantization unit 103 (step 212). The transformed / quantized data is subjected to inverse quantization / inverse transform in the inverse quantization / inverse transform unit 105 (step 213), and is written in the frame memory 106 as a reconstructed image (step 214). Further, the transformed / quantized data is subjected to variable length coding processing in the variable length coding unit 104 (step 215) and output as a bit stream 107. Thereafter, it is determined whether or not to end the encoding process (step 216). If not, the process returns to image input (step 204), and if it ends, the process ends.

  Next, the influence picture number information table 109 in the image coding apparatus according to the present embodiment will be described. FIG. 3 is a diagram showing an outline of the affected picture number information table 109 in the image coding apparatus according to the present embodiment. The number of bits that each element of this table has is defined according to the maximum number of affected pictures. In this table, a table having the same number of elements as the number of pixels of the encoding target image is prepared for the number of reference pictures + 1 or more. 1 is added to the table that stores information on the number of affected pictures for the picture currently being encoded.

  The table of the picture currently being encoded is “0”, the previous picture table is “1”, and the previous two picture table is “2”. As shown in FIG. 3, the influence picture number information is stored in a position corresponding to each pixel. For example, the influence picture number information for the upper left pixel is stored in the upper left position (301). .

  Next, the influence picture number Ne in the image coding apparatus according to the present embodiment will be described. FIG. 4 is a conceptual diagram in which pictures are arranged in the order of processing at the time of image coding in the image coding apparatus according to the present embodiment. The picture currently being encoded is a picture 401, and the picture number is n. The picture processed immediately before is the picture 402, and the picture number is n-1. In this way, it is assumed that picture numbers are assigned according to the order in which encoding processing is performed.

  The affected picture number Ne is an index indicating to which past picture the corresponding area has been affected. For example, the affected picture number Ne of the area affected by the picture number n-1 (picture 402) is 1. In addition, the affected picture number Ne of the area affected by the picture number n-4 (picture 405) to the picture number n (picture 401) is 4. Note that the affected picture number Ne of the area affected only by the picture number n (picture 401) is zero.

  Next, the setting of the image output start picture number Ns111 in the image coding apparatus according to the present embodiment will be described. FIG. 5 is a flowchart at the time of setting the image output start picture number Ns111 in the image coding apparatus according to the present embodiment. When the image output start picture number Ns111 is input (step 501), the image output start picture number Ns111 is stored in the Ns storage area 110 (step 502), and the process ends.

  Here, for the Ns storage area 110, it is conceivable to use a memory, a register, or the like. Further, the image output start picture number Ns111 may be input and set by the user from the outside by an operation of an image processing device in which the image encoding device is mounted, or may be calculated and set by some method. Further, it may be written in advance in the Ns storage area 110 at the time of construction of the image encoding device, and in this case, this operation becomes unnecessary.

  Next, initialization of the affected picture number information table 109 in the image coding apparatus according to the present embodiment will be described. FIG. 6 is a flowchart showing the initialization operation of the affected picture number information table 109 in the image coding apparatus according to the present embodiment. First, the image output start picture number Ns111 is acquired from the Ns storage area 110 (step 601). Next, N is written in the corresponding area of this table, for example, the position of 301 in FIG. 3, using N which is a numerical value larger than the image output start picture number Ns111, for example, N = Ns + 1 (step 602). Next, it is determined whether writing to all elements in all tables is completed (step 603). If writing has not ended, the process returns to step 602 to continue writing, and if it has ended, the process ends.

  Next, the operation of the affected picture number control unit 108 when the prediction process is performed in the image coding apparatus according to the present embodiment will be described. FIG. 7 is a flowchart showing the operation of the affected picture number control unit 108 when the prediction process is executed in the image coding apparatus according to the present embodiment. Here, a case where the immediately preceding picture is referred to in inter-screen prediction will be described.

  First, the pixel position information and prediction mode of the prediction region are acquired from the prediction unit 102 (step 701). The prediction mode includes information indicating whether prediction is performed by intra prediction or inter prediction, and information indicating which picture is referred to in the case of inter prediction. In the image coding apparatus according to the present embodiment, the predicted picture is the previous picture. Next, the prediction mode is determined (step 702). If the prediction is inter-picture prediction, the table “1” of the affected picture number information table 109 is selected (step 704). "Is selected (step 703).

  Next, based on the acquired pixel position information of the prediction area, the number Ne of the affected pictures at the corresponding pixel position in the selected table is acquired (step 705). Thereafter, the prediction mode is determined (step 706), and in the case of inter-screen prediction, 1 is added to the number Ne of affected pictures (step 707). Here, it is determined whether or not the processing of all pixel positions in the prediction region has been completed (step 708). If not, the process returns to determination of the prediction mode for the next pixel position (step 702).

  When the processing of all pixel positions is completed, the maximum value Ne_max of the affected picture number Ne is calculated (step 709), the image output start picture number Ns111 is acquired (step 710), Ne_max and the image output start picture number Ns111. (Step 711). If Ne_max <Ns, an OK signal is output (step 712), and if Ne_max <Ns is not satisfied, an NG signal is output (step 713), and the process ends.

  Next, the operation of the affected picture number control unit 108 at the time of updating the affected picture number information table 109 in the image coding apparatus according to the present embodiment will be described. FIG. 8 is a flowchart showing the operation of the affected picture number control unit 108 when the affected picture number information table 109 is updated in the image coding apparatus according to the present embodiment. Here, a case where the immediately preceding picture is referred to in inter-screen prediction will be described.

  First, the pixel position information and prediction mode of the prediction region are acquired from the prediction unit 102 (step 801). The prediction mode includes information indicating whether prediction is performed by intra prediction or inter prediction, and information indicating which picture is referred to in the case of inter prediction. In the image coding apparatus according to the present embodiment, the predicted picture is the previous picture. Next, the prediction mode is determined (step 802). If the prediction is inter-screen prediction, the “1” table of the affected picture number information table 109 is selected (step 804). "Is selected (step 803).

  Next, based on the acquired pixel position information of the prediction area, the number Ne of the affected pictures at the corresponding pixel position in the selected table is acquired (step 805). Thereafter, the prediction mode is determined (step 806), and in the case of inter-screen prediction, 1 is added to the number of affected pictures Ne (step 807). Thereafter, the number of affected pictures Ne as a result of processing is written in the corresponding processing pixel position in the table “0” of the number-of-affected pictures information table 109 (step 808). Next, it is determined whether or not the processing of all pixel positions in the prediction region has been completed (step 809). If not completed, the process returns to determination of the prediction mode for the next pixel position (step 802), and if completed, the process ends.

  As described above, in the image coding apparatus according to the present embodiment, the image output start picture number Ns111 is set, and the affected picture number Ne that is actually affected by the corresponding picture during the prediction process is calculated in units of pixels. Thus, the control is performed so that the number of affected pictures Ne is smaller than the picture output start picture number Ns111. As a result, fluctuations in the amount of data generated at the time of image encoding can be suppressed, and it is possible to encode randomly accessible images with low delay and high compression without the need for periodic I-picture insertion.

  In this embodiment, only the case where the immediately preceding picture is used as the reference picture at the time of inter-screen prediction has been described. However, the picture used for prediction need not be limited to the immediately preceding picture. A plurality of reference pictures including both directions may be used like multi-reference in H.264 / AVC.

  Also, in the present embodiment, the processing on the image encoding device side for the image output start picture number Ns111 has been described, but the image output start picture number Ns111 is stored in the bitstream 107 and transmitted to the image decoding device side. Also good. In this case, the image decoding apparatus can adjust the image output timing using the image output start picture number Ns111.

<Embodiment 2>
Hereinafter, an image coding apparatus (encoder) according to the second embodiment of the present invention will be described. Since the configuration of the processing module of the image encoding apparatus is the same as the configuration of FIG. 1 in the first embodiment, the description thereof is omitted here.

  First, the affected picture number information table 109 in the image coding apparatus according to the present embodiment will be described. FIG. 9 is a diagram showing an outline of the affected picture number information table 109 in the image coding apparatus according to the present embodiment. The number of bits that each element of this table has is defined according to the maximum number of affected pictures.

  In the affected picture number information table 109 in the image coding apparatus according to the present embodiment, unlike the first embodiment, the affected picture number information is stored in units of blocks, for example, when 4 pixels × 4 pixels are one block. Has one table element in units of 4 pixels × 4 pixels. Accordingly, a table having the same number of elements as the number of blocks of the encoding target picture is prepared for the number of reference pictures + 1 or more. 1 is added to the table that stores the information on the number of affected pictures for the picture that is currently being encoded.

  The table of the picture currently being encoded is “0”, the previous picture table is “1”, and the previous two picture table is “2”. As shown in FIG. 9, the influence picture number information is stored at a position corresponding to each block. For example, when 4 pixels × 4 pixels are used as a block, the influence picture for the upper left 4 pixels × 4 pixels is stored. The number information is stored in the upper left position (901).

  Next, determination of the number of affected pictures Ne in the image coding apparatus according to the present embodiment will be described. In the image coding apparatus according to the present embodiment, the following processing is necessary to handle the number of affected pictures Ne in units of blocks instead of pixels.

  FIG. 10 is a conceptual diagram of a picture used for reference in the image coding apparatus according to the present embodiment. The reference picture 1001 is divided into blocks each having the same number of pixels as the block that is a unit for managing the number of affected pictures Ne. Blocks 1003 to 1006 show some of them. For example, it is assumed that the region referred to by the target region currently being encoded is the region 1002 in FIG. The area 1002 is affected by the blocks 1003 to 1006 as shown in FIG. In order to perform management in units of blocks, the number of affected pictures Ne needs to be set to one value for each block. Therefore, the maximum value of the affected picture number Ne of the related blocks (blocks 1003 to 1006) is selected and used as the affected picture number Ne for the region 1002.

  Next, the operation of the affected picture number control unit 108 when the prediction process is performed in the image coding apparatus according to the present embodiment will be described. FIG. 11 is a flowchart showing the operation of the affected picture number control unit 108 when the prediction process is executed in the image coding apparatus according to the present embodiment. Here, a case where the immediately preceding picture is referred to in inter-screen prediction will be described.

  First, the pixel position information and prediction mode of the prediction region are acquired from the prediction unit 102 (step 1101). The prediction mode includes information indicating whether prediction is performed by intra prediction or inter prediction, and information indicating which picture is referred to in the case of inter prediction. In the image coding apparatus according to the present embodiment, the predicted picture is the previous picture. Next, the prediction mode is determined (step 1102). If the prediction is inter-picture prediction, the table “1” of the affected picture number information table 109 is selected (step 1104). If the prediction is not inter-picture prediction, “0” is selected. "Is selected (step 1103).

  Next, the position of the corresponding block in the prediction area is calculated from the obtained pixel position information of the prediction area (step 1105), and the number of influence pictures Ne of the corresponding block position in the selected table is acquired based on the block position (step 1105). Step 1106). Thereafter, the prediction mode is determined (step 1107), and 1 is added to the number of affected pictures Ne during inter-screen prediction (step 1108). Here, it is confirmed whether the processing of all the pixel positions in the prediction region has been completed (step 1109). If not, the process returns to the determination of the prediction mode for the next pixel position (step 1102).

  When the processing for all pixel positions is completed, the maximum value Ne_max of the affected picture number Ne is calculated (step 1110), the image output start picture number Ns111 is obtained (step 1111), Ne_max and the image output start picture number Ns111 are obtained. (Step 1112). If Ne_max <Ns, an OK signal is output (step 1113). If Ne_max <Ns is not satisfied, an NG signal is output (step 1114), and the process is terminated.

  Next, the operation of the affected picture number control unit 108 at the time of updating the affected picture number information table 109 in the image coding apparatus according to the present embodiment will be described. FIG. 12 is a flowchart showing the operation of the affected picture number control unit 108 when the affected picture number information table 109 is updated in the image coding apparatus according to the present embodiment. Here, a case where the immediately preceding picture is referred to in inter-screen prediction will be described.

  First, the pixel position information and prediction mode of the prediction region are acquired from the prediction unit 102 (step 1201). The prediction mode includes information indicating whether prediction is performed by intra prediction or inter prediction, and information indicating which picture is referred to in the case of inter prediction. In the image coding apparatus according to the present embodiment, the predicted picture is the previous picture. Next, the prediction mode is determined (step 1202). If the prediction is inter-screen prediction, the table “1” of the affected picture number information table 109 is selected (step 1204). If the prediction is not inter-screen prediction, “0” is selected. "Is selected (step 1203).

  Next, the position of the corresponding block in the prediction region is calculated from the obtained pixel position information of the prediction region (step 1205), and the number of affected pictures Ne of the corresponding block position in the selected table is acquired based on the block position (step 1205). Step 1206). Thereafter, the prediction mode is determined (step 1207), and in the case of inter-screen prediction, 1 is added to the number of affected pictures Ne (step 1208). Thereafter, the number of affected pictures Ne as a result of processing is temporarily stored (step 1209). At this time, the maximum number of the affected picture number Ne in the prediction region is finally retained by overwriting and saving only when the affected picture number Ne of the processing result is larger than the already-stored affected picture number Ne. Can be.

  Here, it is determined whether or not the processing of all the pixel positions in the prediction region has been completed (step 1210). If not, the process returns to the determination of the prediction mode for the next pixel position (step 1202). If processing for all pixel positions has been completed, the maximum value of the number of affected pictures Ne temporarily stored in step 1209 is determined as the corresponding processing block in the table “0” of the number of affected picture information table 109. The position is written (step 1211), and the process is terminated.

  In the image coding apparatus according to the present embodiment, when obtaining the maximum value of the number of affected pictures Ne in the prediction area, the corresponding block position is calculated for each pixel in the prediction area, and the number of affected pictures Ne at the block position is calculated. Is obtained, and the maximum value at all pixel positions in the prediction region is obtained. However, this method is an example, and the present invention is not limited to this.

  In this way, by managing the affected picture number information in units of blocks, the prediction accuracy is inferior to the management of the affected picture number information in units of pixels in the first embodiment, but the affected picture number information table The capacity of 109 can be reduced.

  As described above, in the image coding apparatus according to the present embodiment, the image output start picture number Ns111 is set, and the number of affected pictures Ne that is actually affected by the corresponding picture during prediction processing is calculated in units of blocks. Thus, the control is performed so that the number of affected pictures Ne is smaller than the picture output start picture number Ns111. As a result, fluctuations in the amount of data generated at the time of image encoding can be suppressed, and it is possible to encode randomly accessible images with low delay and high compression without the need for periodic I-picture insertion.

  As in the first embodiment, the present embodiment has described only the case where the immediately preceding picture is used as the reference picture at the time of inter-screen prediction. However, the picture used for prediction is not necessarily limited to the immediately preceding picture. H. A plurality of reference pictures including both directions may be used like multi-reference in H.264 / AVC.

  Further, as in the first embodiment, the processing on the image encoding apparatus side for the image output start picture number Ns111 has been described in the present embodiment, but the image output start picture number Ns111 is stored in the bitstream 107 and image decoding is performed. It may be transmitted to the computer apparatus side. In this case, the image decoding apparatus can adjust the image output timing using the image output start picture number Ns111.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  INDUSTRIAL APPLICABILITY The present invention can be used for a method of encoding / decoding a moving image that can be accessed randomly with a small delay time and high compression. Is available.

DESCRIPTION OF SYMBOLS 101 ... Input image, 102 ... Prediction part, 103 ... Transformation / quantization part, 104 ... Variable length coding part, 105 ... Dequantization / inverse transformation part, 106 ... Frame memory, 107 ... Bit stream, 108 ... Influence picture Number control unit, 109 ... Influence picture number information table, 110 ... Ns storage area, 111 ... Image output start picture number Ns,
301... Influence picture number information storage location,
901 ... Influence picture number information storage location,
1001... Reference picture, 1002... Reference region, 1003 to 1006.

Claims (7)

Input image data, and for each predetermined block, at least intra-picture prediction encoding, inter-picture prediction of the image data using a reference picture for the image data and a difference between the reference picture and the image data An image encoding method in an image encoding device that selectively performs encoding,
The image encoding device includes:
A procedure for setting a maximum value of the number of past pictures required for decoding and outputting the image data encoded by the image encoding method as an image output enable parameter;
When the image data is encoded, the number of past pictures required for decoding each pixel of the block to be encoded is calculated for the prediction area for the block to be encoded in units of the block. Obtaining the number of affected pictures to be indicated, and calculating the maximum value of the number of affected pictures in the prediction region as the influence range information;
When the inter-screen predictive coding is selected, if the calculated influence range information exceeds the image output enable parameter, the prediction area is set as another prediction area, and the calculation is performed in any of the prediction areas. When the affected range information exceeds the image output possible parameter, the intra-screen predictive coding is selected to control the range of the reference picture at the time of coding, and
In the calculating procedure, when obtaining the number of affected pictures, a picture number assigned to a picture according to an order in which encoding processing is executed, the picture number of the image data to be encoded, and the image data A method of encoding an image, comprising: obtaining the number of affected pictures based on a difference between a picture number of an oldest picture necessary for decoding the block to be encoded in (1). The image encoding method according to claim 1, wherein:
A table for storing the number of affected pictures in each block having the same number of elements as the number of blocks in the image data to be encoded is prepared for each of the picture currently being encoded and the reference picture, A procedure for writing and initializing the predetermined number of affected pictures in all elements of each table;
For each prediction region, the number of affected pictures is acquired from the element corresponding to the block to which each pixel position of the prediction region belongs in the table corresponding to the prediction region, and the maximum value of the obtained number of affected pictures is obtained. A step of selecting, adding the difference, and writing to the element of the block position to be subjected to predictive encoding processing by the prediction region of the table in the picture currently being encoded. ,
In the calculating step, for each prediction region, the number of affected pictures is obtained from an element corresponding to each block position of the prediction region in the table corresponding to the prediction region. . The image encoding method according to claim 2, wherein
When the intra-prediction coding is selected, the control procedure is performed when each pixel position of the prediction area of the table in the picture currently being subjected to the coding process is used for the prediction. An image coding method characterized in that the number of affected pictures is respectively acquired from an element corresponding to a block to which the image belongs, and the maximum value of the obtained number of affected pictures is used as the affected range information. The image encoding method according to claim 1, wherein:
The image encoding apparatus stores the image output enable parameter in a bitstream encoded and output by the image encoding method, and transmits the parameter to the image decoding apparatus. Input image data, and for each predetermined block, at least intra-picture prediction encoding, inter-picture prediction of the image data using a reference picture for the image data and a difference between the reference picture and the image data An image encoding device that selectively performs encoding,
Means for setting, as an image output possible parameter, a maximum value of the number of past pictures required to decode and output the image data encoded by the image encoding method;
When the image data is encoded, the number of past pictures required for decoding each pixel of the block to be encoded is calculated for the prediction area for the block to be encoded in units of the block. Means for obtaining the number of affected pictures shown, and calculating the maximum value of the number of affected pictures in the prediction region as the influence range information;
When the inter-screen predictive coding is selected, if the calculated influence range information exceeds the image output enable parameter, the prediction area is set as another prediction area, and the calculation is performed in any of the prediction areas. Means for controlling the range of the reference picture at the time of encoding by selecting intra prediction encoding when the affected range information exceeds the image output enable parameter,
In the calculating means, when obtaining the number of affected pictures, a picture number assigned to a picture in accordance with an order in which encoding processing is executed, the picture number of the image data to be encoded, and the image data An image coding apparatus characterized in that the number of affected pictures is obtained from a difference between a picture number of an oldest picture necessary for decoding the block to be coded in (1).   An image processing apparatus, wherein the image encoding device according to claim 5 is mounted. Input image data, and for each predetermined block, at least intra-picture prediction encoding, inter-picture prediction of the image data using a reference picture for the image data and a difference between the reference picture and the image data An image encoding method in an image encoding device that selectively performs encoding,
The image encoding device includes:
A procedure for setting a maximum value of the number of past pictures required for decoding and outputting the image data encoded by the image encoding method as an image output enable parameter;
And a step of storing the image output enable parameter in a bit stream encoded and output by the image encoding method and transmitting the parameter to an image decoding apparatus.

Images