WO2009128208A1

WO2009128208A1 - Dynamic image encoder, dynamic image decoder, dynamic image encoding method, and dynamic image decoding method

Info

Publication number: WO2009128208A1
Application number: PCT/JP2009/001449
Authority: WO
Inventors: 高橋昌史; 山口宗明; 伊藤浩朗
Original assignee: 株式会社日立製作所
Priority date: 2008-04-16
Filing date: 2009-03-30
Publication date: 2009-10-22
Also published as: JPWO2009128208A1

Abstract

Provided is a dynamic image encoder comprised of an interframe prediction unit (106) which performs the interframe prediction and calculates the prediction error, a motion description unit (111) which models the motion information of a target region crossing a plurality of frames, a frequency converter (109) and a quantization processor (110) which encode the prediction error, and a variable-length encoder (113) which performs the variable-length coding corresponding to the symbol probability based on the information modeled by the motion description unit (111). The motion description unit (111) specifies the start frame and the end frame in a plurality of frames, and models the motion of the target region between the start frame and the end frame by a temporal function.

Description

Moving picture encoding apparatus, moving picture decoding apparatus, moving picture encoding method, and moving picture decoding method

The present invention relates to a moving picture encoding technique for encoding a moving picture and a moving picture decoding technique for decoding a moving picture.

さまざま Various standards have been established as international standard encoding methods for recording and transmitting large volumes of moving image information as digital data.

Some of these encoding methods have been adopted as encoding methods for digital satellite broadcasting, DVDs, mobile phones, digital cameras, and the like, and the range of use is now expanding and becoming familiar.

In these standards, the encoding target image is predicted in block units using the image information that has been encoded, and the prediction difference from the original image is encoded, thereby eliminating the redundancy of the moving image. The code amount is reduced.

In particular, in inter-screen prediction that refers to an image different from the target image, high-precision prediction is enabled by searching the reference image for a block having a high correlation with the encoding target block.

However, in the conventional inter-screen prediction, it is necessary to encode the result of the block search as a motion vector in addition to the prediction difference, resulting in a code amount overhead.

H. is one of the standards for such inter-screen prediction. The H.264 / AVC (Advanced Video Coding) standard introduces a prediction technique for motion vectors in order to reduce the amount of motion vector codes.

In other words, when encoding a motion vector, the motion vector of the target block is predicted using an encoded block located around the target block, and the difference between the prediction vector and the motion vector (difference vector) is variable. Encode long.

This has succeeded in significantly reducing the amount of motion vector codes. The accuracy of motion vector prediction by H.264 / AVC is not sufficient, and a large amount of code is still required for motion vectors, especially for images with complex motion such as the presence of multiple moving objects. was there.

Accordingly, an object of the present invention is to improve a compression efficiency by reducing a code amount of a motion vector by improving a difference vector calculation method, a moving image decoding device, a moving image It is to provide an encoding method and a moving image decoding method.

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 this application, the outline of typical ones will be briefly described as follows.

That is, the outline of a typical one is that a moving image coding apparatus specifies a start frame and an end frame within a plurality of frames, and models the movement of a target region between the start frame and the end frame by a time function. It has a motion description part.

The moving image decoding apparatus further includes a variable length decoding unit that decodes a motion vector of the target region based on a time function in which the motion of the target region between the start frame and the end frame is modeled. Is.

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

In other words, the effect obtained by a typical one is that by improving the calculation method of the difference vector, the code amount of the motion vector can be reduced to improve the compression efficiency, and a high-quality video with a small code amount can be obtained. Can be provided.

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.

First, an inter-screen prediction process that is a premise of a moving picture coding method and a moving picture decoding method according to an embodiment of the present invention will be described with reference to FIGS. FIG. 1 is an explanatory diagram for explaining an inter-screen prediction process that is a premise of a moving picture coding method and a moving picture decoding method according to an embodiment of the present invention. 2 conceptually shows the operation of the inter-screen prediction processing by H.264 / AVC. FIG. 2 is an explanatory diagram for explaining calculation of a prediction vector in inter-frame prediction processing which is a premise of the video encoding method and video decoding method according to an embodiment of the present invention. This is a conceptual illustration of how to do this.

H. In H.264 / AVC, the encoding target image is encoded in block units according to the raster scan order.

When performing inter-screen prediction, as shown in FIG. 1, a decoded image of an encoded image included in the same video 501 as the encoding target image 503 is used as a reference image 502, and the target block 504 in the target image A block 505 having a high correlation is searched from the reference image.

At this time, in addition to the prediction difference calculated as the difference between both blocks, the difference between the coordinate values of both blocks is encoded as a motion vector 506. On the other hand, the reverse procedure described above may be performed at the time of decoding, and the decoded image can be acquired by adding the decoded prediction difference to the block 505 in the reference image.

H. In H.264 / AVC, in order to reduce the overhead of the code amount due to the motion vector described above, a prediction technique for the motion vector is introduced. That is, when a motion vector is encoded, a motion vector of the target block is predicted using an encoded block located around the target block, and a difference vector between the prediction vector and the motion vector is encoded. At this time, since the size of the difference vector is concentrated to almost zero, the amount of codes can be reduced by variable-length encoding the difference vector.

As shown in FIG. 2, the prediction vector is calculated by setting the encoded blocks adjacent to the left side, the upper side, and the upper right side of the target block 601 as the block A 602, the block B 603, and the block C 604, respectively. MVA, MVB, and MVC.

At this time, the prediction vector PMV is calculated like the prediction vector PMV605 using a function Median that returns the median value of a plurality of values designated as arguments. Further, the difference vector DMV is calculated as the difference vector 606 between the motion vector MV of the target block and the prediction vector PMV, and then the DMV is variable-length encoded.

As mentioned above, H. In H.264 / AVC, it has become possible to significantly reduce the amount of code required for motion vectors by introducing a prediction technique for motion vectors. However, H. In the case of H.264 / AVC, only the neighboring blocks in the spatial direction are considered when calculating the prediction vector, and it cannot be said that the motion of the object is necessarily reflected.

For this reason, the motion vector prediction accuracy is not sufficient particularly in an image having a plurality of moving objects, and a large amount of code is still required for the motion vector.

In this embodiment, as will be described later, the prediction accuracy for a motion vector can be improved by modeling the motion of the encoding target region as a time function and using it for calculation of a prediction vector.

Next, modeling of a moving picture coding method and a moving picture decoding method according to an embodiment of the present invention will be described with reference to FIGS. FIG. 3 is a diagram showing an example of motion modeling of the moving picture encoding method and the moving picture decoding method according to the embodiment of the present invention, and FIG. 4 is a moving picture encoding according to the embodiment of the present invention. FIG. 5 is a diagram showing an example of motion vector encoding of the method and the video decoding method, and FIG. 5 is a selection of start and end frames of the video encoding method and video decoding method according to an embodiment of the present invention. FIG. 6 is a diagram illustrating another example of motion modeling of the moving picture coding method and the moving picture decoding method according to an embodiment of the present invention, and FIG. 7 is a diagram illustrating one embodiment of the present invention. It is a figure which shows the other example of selection of the start frame and end frame of the moving image encoding method and moving image decoding method which concern on this form.

As shown in FIG. 3, in the present embodiment, first, a start frame 701 and an end frame 705 are prepared with n−1 frames 702 to 704 interposed therebetween. Subsequently, an area 707 corresponding to the specific area 708 in the end frame is searched in the start frame, and the movement of the target area is modeled as a time function MVMt 706 based on the difference value of the coordinates of these corresponding areas.

Using this motion model, the motion vectors in the frames 702 to 704 sandwiched between the start frame and the end frame are encoded. Further, the modeled motion information is separately encoded and stored in a stream.

For example, in the example shown in FIG. 3, the movement of the target area is modeled linearly and represented by a linear function at time t. In this case, the coefficients A, B, C, and D of the function MVMt 706 are encoded as motion parameters.

Furthermore, for example, information for specifying a range of a frame to which motion modeling by the function MVMt can be applied, such as a start frame number and an end frame number, and region information to which motion modeling can be applied, such as a block number Is encoded.

At this time, there is no particular question as to how to model the movement of the target area. That is, in the example shown in FIG. 3, the movement of the target region is approximated by a straight line, but this is approximated by, for example, an ellipse, a quadratic parabola, a Bezier curve, a clothoid curve, a cycloid, a reflection, a pendulum motion, etc. It doesn't matter.

In FIG. 4, a motion vector is encoded using a motion model represented by the function MVMt. Here, a frame 801 with a start frame time t = 0 and an end frame time t = n. A case where a frame 802 with frame time t = m sandwiched between frames 803 is encoded is shown.

As already described, when the motion vector MV804 of the target block 808 is encoded, the motion MVMt 805 modeled using the region 807 in the start frame and the region 809 in the corresponding end frame is used.

That is, the motion MVMm at time t = m is set as the predicted value of the motion vector MV, and the difference vector DMV806 is calculated from the difference between the motion vector MV and the predicted vector MVMm. Further, H.C. As with H.264 / AVC, DMV is variable length encoded.

In the above, the motion vector encoding method in the frame sandwiched between the start frame and the end frame has been described, but the motion vector encoding method in the start frame and the end frame is not particularly limited. However, for example, as shown in FIG. Similar to the H.264 / AVC method, it is effective to use a method of predicting a motion vector by referring to the peripheral blocks of the target block.

Further, the selection method of the start frame and the end frame is not particularly limited. For example, in the example illustrated in FIG. 5, inter-picture prediction using an I picture 901 that can use only intra-prediction and one reference image. In this example, encoding is performed using

P pictures

904 and 907 that can be encoded and

B pictures

902, 903, 905, and 906 that are capable of inter-screen prediction using two reference images.

In this case, after the I picture 901 is encoded, the P picture 904 is encoded with reference to this. Subsequently, the B pictures 902 and 903 are encoded with reference to the two encoded

images

901 and 904. Similarly, the P picture 907 is encoded next, and then the B pictures 905 and 906 are encoded with reference to the two

images

904 and 907.

In such a picture structure, it is effective to apply this embodiment as follows, for example. That is, first, the I picture 901 and the P picture 904 are encoded, and the motion is modeled using the two images as a start frame and an end frame, respectively.

Then, B pictures 902 and 903 in between are encoded using this model. Subsequently, the next P picture 907 is encoded, and motion modeling is performed using the two

P pictures

904 and 907 as a start frame and an end frame, respectively. Then, B pictures 905 and 906 in between are encoded using this model.

By selecting the start frame and end frame as described above, Compared with H.264 / AVC, encoding can be performed without causing a particularly large delay.

Also, as shown in FIG. 6, encoding using a motion model may be combined with encoding not using a motion model. Here, the first frame 1301 is set as the start frame, and the end frame is selected according to the nature of the image.

In this example, an end frame 1305 is specified with three

frames

1302, 1303, and 1304 sandwiched between the start frame and the start frame. The method for determining the end frame is not particularly limited. For example, the range of the frame in which the motion of the target region can be represented by the motion model MVMt 706 in FIG. 3 is specified, and the end frame is sandwiched with the start frame It is effective to determine the frame.

Further, when the motion of the target region cannot be modeled by the motion model MVMt 706 of FIG. 3 or the like, for example, the conventional method as shown in FIG. The motion vector may be encoded by By doing so, encoding can be performed efficiently.

FIG. 7 shows another example of the method for selecting the start frame and the end frame. Here, the type of picture is not particularly limited, but for simplicity, only the case of using only an I picture and a P picture is shown.

In this case, encoding is performed in the same order as the video display order (1001 → 1002 → 1003 →...). In this example, first, an I picture 1001 is encoded, and this is used as a start frame, and an image 1004 after n−1 frames is used as an end frame to model motion.

At this time, for the image 1004, only the corresponding region search for modeling the motion is performed, and the encoding process is not performed. Subsequently, the image (1002, 1003,...) Is encoded using this model.

In this case, since it is necessary to pre-read the image 1004 after the image 1001 is encoded, a large delay occurs when encoding, but many images are sandwiched between the start frame and the end frame. There is an advantage that the efficiency of motion modeling can be increased.

In each of the above examples, the motion modeling is performed in units of blocks, but in addition, for example, the modeling may be performed in units of objects separated from the background of the image.

Next, the configuration and operation of the moving picture coding apparatus according to the embodiment of the present invention will be described with reference to FIG. FIG. 8 is a block diagram showing the configuration of the moving picture coding apparatus according to the embodiment of the present invention.

In FIG. 8, the moving image coding apparatus includes an input image memory 102 that holds an input original image 101, a block dividing unit 103 that divides the input image into small regions, and an intra-screen prediction unit that performs intra-screen prediction in units of blocks. 105. Determine an inter-screen prediction unit 106 that performs inter-screen prediction on a block basis based on the amount of motion detected by the motion search unit 104, and predictive encoding means (prediction method and block size) that matches the nature of the image. A mode selection unit 107, a subtraction unit 108 for generating a prediction difference, a frequency conversion unit 109 and a quantization processing unit 110 for encoding the prediction difference, a motion description unit 111 for modeling a motion in the target region, A motion information memory 112 that holds modeled motion information, a variable length encoding unit 113 that performs encoding according to the probability of occurrence of symbols, and encoding once An inverse quantization processing unit 114 and an inverse frequency transform unit 115 for decoding the predicted difference, an addition unit 116 for generating a decoded image using the decoded prediction difference, and holding the decoded image The reference image memory 117 is used for later prediction.

The input image memory 102 holds one image from the original image 101 as an encoding target image, and divides the image into fine blocks by the block dividing unit 103, and a motion search unit 104 and an in-screen prediction unit 105. , And the inter-screen prediction unit 106.

The motion search unit 104 calculates the motion amount of the corresponding block using the decoded image stored in the reference image memory 117, and passes the motion vector to the inter-screen prediction unit 106. The intra-screen prediction unit 105 and the inter-screen prediction unit 106 execute the intra-screen prediction process and the inter-screen prediction process in units of blocks of several sizes, and the mode selection unit 107 selects an optimal prediction method.

Subsequently, the subtraction unit 108 generates a prediction difference by the optimal prediction encoding means (prediction method and block size) and passes it to the frequency conversion unit 109. The frequency conversion unit 109 and the quantization processing unit 110 perform frequency conversion and quantization processing such as DCT (Discrete Cosine Transformation) in units of blocks having a size specified for the transmitted prediction difference. And pass it to the variable length coding unit 113 and the inverse quantization processing unit 114.

Further, the motion description unit 111 models the motion of the target region by a time function based on information (image information, motion vector, etc.) regarding the start frame and the end frame, and information such as a start frame number, an end frame number, and a motion parameter. Is sent to the motion information memory 112 for storage.

Furthermore, in the variable length coding unit 113, the prediction difference information represented by the frequency conversion coefficient, for example, the prediction direction used when performing the intra prediction, the motion vector used when performing the inter prediction, and the motion Information necessary for decoding, such as motion parameters used for modeling, is subjected to variable length coding based on the probability of symbol generation to generate a coded stream.

At this time, a motion model stored in the motion information memory 112 is used to encode motion vectors in frames other than the start frame and the end frame. In addition, the inverse quantization processing unit 114 and the inverse frequency transform unit 115 perform inverse frequency transform such as inverse quantization and IDCT (Inverse DCT) on the quantized frequency transform coefficients, respectively, so that the prediction difference Is sent to the adder 116. Subsequently, the adder 116 generates a decoded image and stores it in the reference image memory 117.

Next, the configuration and operation of the moving picture decoding apparatus according to an embodiment of the present invention will be described with reference to FIG. FIG. 9 is a block diagram showing the configuration of the video decoding apparatus according to one embodiment of the present invention.

In FIG. 9, the moving picture decoding apparatus includes, for example, a variable length decoding unit 202 that performs the reverse procedure of variable length coding on the coded stream 201 generated by the moving picture coding apparatus illustrated in FIG. 8. An inverse quantization processing unit 203 and an inverse frequency transform unit 204 for decoding the prediction difference, a motion information memory 205 for storing information necessary for motion modeling such as motion parameters, start frame numbers, and end frame numbers; An inter-screen prediction unit 206 that performs inter-screen prediction, an intra-screen prediction unit 207 that performs intra-screen prediction, an adder unit 208 for acquiring a decoded image, and a reference image memory for temporarily storing the decoded image 209.

The variable length decoding unit 202 performs variable length decoding on the encoded stream 201 and acquires information necessary for prediction processing such as a frequency transform coefficient component of a prediction difference, a block size, a motion vector, and a motion parameter.

The former prediction difference information is sent to the inverse quantization processing unit 203, and the information necessary for the latter prediction processing is sent to the motion information memory 205, the inter-screen prediction unit 206, or the intra-screen prediction unit 207. .

Subsequently, the inverse quantization processing unit 203 and the inverse frequency transform unit 204 perform decoding by performing inverse quantization and inverse frequency transform on the prediction difference information, respectively. Further, the motion information memory 205 stores information necessary for motion modeling such as motion parameters.

Subsequently, the inter-screen prediction unit 206 or the intra-screen prediction unit 207 executes a prediction process with reference to the reference image memory 209 based on the information sent from the variable length decoding unit 202, and the addition unit 208 performs decoding. A converted image is generated and the decoded image is stored in the reference image memory 209.

Next, the configuration and operation of the motion description unit of the video encoding device according to the embodiment of the present invention will be described with reference to FIGS. FIG. 10 is a block diagram showing an example of the configuration of a motion picture encoding apparatus motion description section according to an embodiment of the present invention. FIG. 11 is a diagram of the motion picture encoding apparatus motion description section according to an embodiment of the present invention. It is a block diagram which shows the other example of a structure.

In FIG. 10, the motion description unit 111 receives the start frame and the end frame original image or the target image 301, and searches for a corresponding area between the start frame and the end frame, and a start frame memory 302 for storing the start frame. It includes a motion search unit 303 that performs the motion information modeling unit 304 that performs motion modeling based on the search result, and outputs motion information such as the motion parameter 305.

First, the original image or decoded image of the start frame is input to the motion description unit 111 and stored in the start frame memory 302. Subsequently, when the original image or decoded image of the end frame is input, the motion search unit 303 searches for the corresponding area between the start frame stored in the start frame memory 302 and the input end frame. The search result is passed to the motion information modeling unit 304.

The motion information modeling unit 304 performs motion modeling based on the search result. In the motion information modeling unit 304, for example, the motion of the target region is modeled by the function MVMt 306 in FIG.

11, a motion description unit 111 receives a motion vector 401 calculated when encoding a start frame and an end frame, a start frame memory 402 for storing the motion vector of the start frame, and a motion vector of the start frame. And a motion information modeling unit 403 that models the motion of the target region from the motion vector of the end frame, and outputs motion information such as the motion parameter 404.

First, a motion vector calculated when a start frame is encoded is input to the motion description unit 111 and stored in the start frame memory 402. Subsequently, when the motion vector calculated when the end frame is encoded is input, the motion information modeling unit 403 performs motion modeling based on both vectors. In the motion information modeling unit 403, for example, the motion of the target region is modeled by the function MVMt405 in FIG.

Next, with reference to FIG. 12, a description will be given of an encoding process procedure for one frame of the moving picture encoding apparatus according to the embodiment of the present invention. FIG. 12 is a flowchart showing a one-frame encoding process procedure of the video encoding apparatus according to the embodiment of the present invention.

First, the following processing is performed as loop 1 for all the blocks present in the frame to be encoded (step 1101). That is, prediction is executed as loop 2 for all combinations of encoding mode prediction methods and block sizes once for the corresponding block (step 1102).

Here, it is determined whether the mode is the intra prediction mode (step 1103), and in accordance with the determination in step 1103, the intra prediction process (step 1104) or the inter prediction process (step 1105) is performed to calculate the prediction difference. Do.

Furthermore, when performing inter-screen prediction, a motion vector is encoded in addition to the prediction difference. Here, it is determined whether the frame is a start frame (step 1106). If the target frame is the start frame in step 1106, the motion vector used for inter-screen prediction is stored by, for example, the method shown in FIG. 1107).

On the other hand, if the target frame is not the start frame in step 1106, it is determined whether it is the end frame (step 1108). If the target frame is the end frame in step 1108, the motion vector of the corresponding area and the stored start frame are determined. The motion is modeled using the motion vector of, and a motion parameter is calculated (step 1109).

In step 1106 and step 1108, if the target frame is neither the start frame nor the end frame, the difference vector DMV is calculated using the motion model (step 1111).

Note that the DMV calculation for the start frame and end frame is H.264. This is performed by a conventional method using H.264 / AVC (step 1110).

Subsequently, frequency conversion processing (step 1112), quantization processing (step 1113), and variable length encoding processing (step 1114) are performed on the prediction difference, and image quality distortion and code amount of each encoding mode are calculated.

If the above processing is completed for all coding modes by loop 2, the mode with the highest coding efficiency is selected based on the above results (step 1115).

When selecting the one with the highest coding efficiency from among a large number of coding modes, for example, the RD-Optimization method that determines the optimum coding mode from the relationship between the image quality distortion and the code amount is used. Thus, encoding can be performed efficiently.

For details of the RD-Optimization method, refer to the following documents.
“G. Sullivan and T. Wiegand:“ Rate-Distration Optimization for Video Compression ”, IEEE Signal Processing Magazine, vol. 15, no. 6, pp. 74-90.

Subsequently, for the selected encoding mode, the quantized frequency transform coefficient is subjected to inverse quantization processing (step 1116) and inverse frequency transform processing (step 1117) to decode the prediction difference, and the decoded image Is stored in the reference image memory (step 1118).

If the above processing is completed for all the blocks by loop 1, the encoding for one frame of the image is completed (step 1119).

Next, with reference to FIG. 13, a description will be given of a one-frame encoding process procedure of the video decoding apparatus according to the embodiment of the present invention. FIG. 13 is a flowchart showing a one-frame encoding process procedure of the moving picture decoding apparatus according to the embodiment of the present invention.

First, the following processing is performed as loop 1 for all blocks in one frame (step 1201). That is, variable length decoding processing is performed on the input stream (step 1202), and inverse quantization processing (step 1203) and inverse frequency conversion processing (step 1204) are performed to decode the prediction difference.

Subsequently, it is determined whether the mode is an intra-screen prediction mode (step 1205), and an intra-screen prediction process (step 1206) or an inter-screen prediction process (step 1210) is performed according to the determination in step 1205.

Note that when performing inter-screen prediction, it is necessary to decode the motion vector MV prior to the prediction. Here, it is determined whether the current frame is a start frame or an end frame (step 1207). The MV is decoded by a conventional method based on H.264 / AVC (step 1208).

On the other hand, if it is determined in step 1207 that the target frame is neither a start frame nor an end frame, MV decoding is performed using the motion model (step 1209).

When the above processing is completed for all the blocks in the frame by loop 1, decoding for one frame of the image is completed (step 1211).

In this embodiment, DCT is cited as an example of frequency conversion, but DST (Discrete Sine Transformation), WT (Wavelet Transformation), DFT (Discrete Fourier Transform: Discrete Fourier Transform). Any orthogonal transform can be used for removing the correlation between pixels, such as KLT (Karhunen-Loeve Transformation), and even if the prediction difference itself is encoded without frequency conversion. I do not care.

Furthermore, there is no need to perform variable length coding. In the present embodiment, the prediction vector is calculated using the function MVMt, but the motion vector itself may be expressed using this function. In this case, the motion vector MV is equal to MVMt, and there is no need to encode the difference vector DMV.

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.

The present invention relates to a moving picture coding technique for coding a moving picture and a moving picture decoding technique for decoding a moving picture, and can be widely applied to apparatuses that perform coding and decoding of a moving picture.

It is explanatory drawing for demonstrating the prediction process between screens used as the premise of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is explanatory drawing for demonstrating calculation of the prediction vector of the inter-screen prediction process used as the premise of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is a figure which shows an example of the motion modeling of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is a figure which shows an example of the encoding of the motion vector of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is a figure which shows an example of selection of the start frame and end frame of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is a figure which shows the other example of the motion modeling of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is a figure which shows the other example of selection of the start frame and the end frame of the moving image encoding method and moving image decoding method which concern on one embodiment of this invention. It is a block diagram which shows the structure of the moving image encoder which concerns on one embodiment of this invention. It is a block diagram which shows the structure of the moving image decoding apparatus which concerns on one embodiment of this invention. It is a block diagram which shows an example of a structure of the moving image encoder motion description part which concerns on one embodiment of this invention. It is a block diagram which shows the other example of a structure of the moving image encoder motion description part which concerns on one embodiment of this invention. It is a flowchart which shows the encoding process procedure of 1 frame of the moving image encoder which concerns on one embodiment of this invention. It is a flowchart which shows the encoding process procedure of 1 frame of the moving image decoding apparatus which concerns on one embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Original image, 102 ... Input image memory, 103 ... Block division part, 104 ... Motion search part, 105 ... In-screen prediction part, 106 ... Inter-screen prediction part, 107 ... Mode selection part, 108 ... Subtraction part, 109 ... Frequency conversion unit, 110 ... quantization processing unit, 111 ... motion description unit, 112 ... motion information memory, 113 ... variable length coding unit, 114 ... inverse quantization processing unit, 115 ... inverse frequency conversion unit, 116 ... addition unit 117: Reference image memory, 201: Encoded stream, 202: Variable length decoding unit, 203 ... Inverse quantization processing unit, 204 ... Inverse frequency transform unit, 205 ... Motion information memory, 206 ... Inter-screen prediction unit, 207 ... intra-screen prediction unit, 208 ... adding unit, 209 ... reference image memory, 301 ... target image, 302 ... start frame memory, 303 ... motion search unit, 304 ... motion information modeling unit 305 ... Motion parameters, 306 ... Motion modeling functions, 401 ... Motion vectors, 402 ... Start frame memory, 403 ... Motion information modeling unit, 404 ... Motion parameters, 405 ... Motion modeling functions, 501 ... Images, 502 Reference image, 503 ... Image to be encoded, 504, 505 ... Block, 601-606 ... Block, 701-705 ... Frame, 706 ... Motion modeling function, 707, 708 ... Region, 801-803 ... Frame, 804 ... motion vector MV, 805 ... motion MVMt, 806 ... difference vector DMV, 807 to 809 ... area, 901 ... I picture, 904, 907 ... P picture, 902, 903, 905, 906 ... B picture, 1001 ... I picture, 1002 to 1004 ... I picture, 1301 to 1307 ... frame

Claims

An inter-screen prediction unit that performs inter-screen prediction and calculates a prediction difference;
A motion description unit that models motion information of a target area across multiple frames;
A frequency conversion unit and a quantization processing unit that perform encoding on the prediction difference;
A variable length coding unit that performs variable length coding according to the probability of occurrence of a symbol based on the modeling information by the motion description unit;
The motion description unit specifies a start frame and an end frame in the plurality of frames, and models the motion of the target area between the start frame and the end frame by a time function. Image encoding device.
The moving picture encoding apparatus according to claim 1,
The variable length coding unit is calculated by using the motion vector used when performing the inter-screen prediction on the target region and the time function obtained by modeling the motion of the target region in the motion description unit. A moving picture coding apparatus characterized in that variable length coding is performed on a difference between vectors.
The moving picture encoding apparatus according to claim 1,
The inter-picture prediction unit performs motion compensation using a vector calculated by the time function obtained by modeling the motion of the target region in the motion description unit.
The moving picture encoding apparatus according to claim 1,
The moving picture coding apparatus, wherein the motion description unit designates the start frame and the end frame within a range that can be modeled by the time function.
An inter-screen prediction unit that performs inter-screen prediction and calculates a prediction difference, a motion description unit that models motion information of a target region over a plurality of frames, a frequency conversion unit that performs encoding on the prediction difference, and a quantum Coding in a moving picture coding apparatus comprising: a coding processing unit; and a variable length coding unit that performs variable length coding according to a symbol generation probability based on the modeling information by the motion description unit Because
The motion description unit designates a start frame and an end frame within the plurality of frames, and the motion of the target region between the start frame and the end frame is modeled by a time function. Video encoding method.
The moving image encoding method according to claim 5, wherein
Calculated by the variable length encoding unit using the motion vector used when performing the inter-screen prediction on the target region and the time function modeling the motion of the target region in the motion description unit. A moving image encoding method, wherein variable-length encoding is performed on a difference between vectors.
The moving image encoding method according to claim 5, wherein
A moving picture coding method, wherein motion compensation is performed by the inter-frame prediction unit using a vector calculated by the time function obtained by modeling the motion of the target region in the motion description unit.
The moving image encoding method according to claim 5, wherein
The moving picture coding method, wherein the motion description unit designates the start frame and the end frame within a range that can be modeled by the time function.
A variable length decoding unit that decodes variable length encoded data in the reverse procedure of the variable length encoding;
An inverse quantization processing unit and an inverse frequency transform unit for decoding the prediction difference;
An inter-screen prediction unit that performs inter-screen prediction and obtains a decoded image,
The variable length decoding unit decodes the motion vector of the target region based on a time function in which the motion of the target region between the start frame and the end frame is modeled. Device.
The moving picture decoding apparatus according to claim 9, wherein
The video decoding apparatus, wherein the inter-screen prediction unit performs motion compensation using a motion vector calculated by a time function that models the motion of the target region.
A variable-length decoding unit that decodes variable-length encoded data in the reverse procedure of the variable-length encoding; an inverse quantization processing unit that decodes a prediction difference; and an inverse frequency conversion unit; A moving picture decoding method in a moving picture decoding apparatus having an inter-screen prediction unit that performs a decoded image and performs:
A moving image in which a motion vector of the target area is decoded based on a time function in which the motion of the target area between a start frame and an end frame is modeled by the variable length decoding unit. Decryption method.
The video decoding method according to claim 11, wherein
A moving picture decoding method, wherein motion compensation is performed by the inter prediction unit using a motion vector calculated by a time function modeling the movement of the target region.