JP2002051342A - Coding apparatus, method of coding and storage medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the coding processing time or motion compensation accuracy by using shape information of an image with contour information when coding this image, and also to enable coding of an image which does not have contour information by extracting the contour information from this image to use the contour information. SOLUTION: The presence/absence of information according to blocked image data to be encoded is confirmed. When the presence/absence of the information is not detected, contour information with respect to the blocked image data is formed and predicted values of motion vectors of the blocked image data are calculated, and a first region which most resembles the blocked image data to be coded is determined. Motion vectors of the blocked image data to be coded are calculated, difference block value blocks are encoded, and the coded difference values are decoded to be re-stored in the storage means.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoding apparatus for encoding image data of each frame of a moving image, a method thereof, and a storage medium.

[0002]

2. Description of the Related Art An example of a conventional encoding apparatus is shown in FIG.
Shown in

In FIG. 7, 701 is a buffer memory for original image data, 702 is a block circuit, 703 is a subtraction circuit, 704 is an orthogonal transformation circuit, 705 is a quantization circuit,
706 is a variable length coding circuit, 707 is an inverse quantization circuit, 7
08 is an inverse orthogonal transformation circuit, 709 is an addition circuit, and 710 and 7
14 and 715 are selector switches, 711 and 712 are frame memories, 713 is an average value calculation circuit, 716 is a control circuit, and 717 is a motion vector search circuit.

In FIG. 7, original image data to be encoded is stored in a buffer memory 701,
16 is read out in response to a request from the
02 is divided into blocks (n × m sizes) corresponding to the coding unit.

[0005] All subsequent processing is performed in this block unit.

The original image data (blocked image data) that has been blocked is input to a vector search circuit 717 to determine a motion vector (motion compensation).

FIG. 8 shows an example of a motion vector search method in the motion vector search circuit 717.

FIG. 8A shows a frame I (t) to be coded at the current time t, and FIG. 8B shows a reference frame I (t-t) corresponding to the frame to be coded shown in FIG.
1) and FIG. 8 (c) show a reference frame I (t-2) for the frame shown in FIG. 8 (b). In FIG. 8A, the shaded block (block A) is the target block in I (t), and MVc is the motion vector of block A in I (t) with respect to block A in I (t-1). . Also, in FIG. 8C, MVp is I
The motion vector of the block A at (t-1) with respect to the block A at I (t-2).

Here, the search method of the motion vector MVc follows the following procedure.

First, it is assumed that the motion vector MVp has been detected. Then, a position shifted by MVp from the same position as the block A on I (t-1) is obtained, and this position is set as the search start origin.

Next, a predetermined size range (motion vector search range) is set in the horizontal and vertical directions around the search start origin, and the frame of the predetermined range is shifted by one pixel within this search range. , And the sum of absolute differences obtained by the following equation (Σ | (pixels of block A of I (t) −pixels within a frame of a predetermined range |)) is calculated. Then, by comparing the sum of absolute differences with one-pixel accuracy, the coordinate value at which the minimum value is obtained (the coordinate value at the upper right position of the frame in the predetermined range in the figure) is detected. In order to further improve the accuracy of motion vector search, virtual pixels (pixel values are calculated from surrounding pixels) are arranged between pixels within the frame of the predetermined range. A virtual double resolution is prepared in the frame of, and the difference absolute value sum for each size of the block A is calculated in this frame in the same manner as described above. As a result, by comparing each of the obtained sums of the absolute values of the differences, and detecting the coordinate value at which the minimum value is obtained,
The motion vector MVc can be obtained.

In the above-described motion vector search method, I (t-1) shown in FIG.
Under the control of 16, each pixel within the above-described predetermined range frame is sequentially read from the frame memory 711 or 712, and is input to the motion vector search circuit 717 via the selector switches 714 and 715 a.

At this time, in the above example, the sum of absolute differences between each pixel in the input frame within the predetermined range and the block A of I (t) is calculated, and as a result, a motion vector is determined.

Next, the block image data is subtracted from the subtraction circuit 7.
03 is input. At the same time, under the control of the control circuit 716, the predicted value (locally decoded block) specified by the motion vector obtained by the motion vector search circuit 717
Is read out from the frame memory 711 or 712, input to the subtraction circuit 703 via the selector switches 714 and 715b, and subtraction processing is performed from the blocked image data from the blocking circuit 702.

The result of the subtraction in the subtraction circuit 703 is output as a prediction error of the data of the original image, and is then orthogonally transformed in the orthogonal transformation circuit 704.

The prediction error orthogonally transformed by the orthogonal transformation circuit 704 is quantized by a quantization circuit 705 and then input to a variable length encoding circuit 706, where a variable length code based on a predetermined function is added. You.

On the other hand, the prediction error after the quantization processing is subjected to processing completely opposite to that of the quantization circuit 705 and the orthogonal transformation circuit 704 by the inverse quantization circuit 707 and the inverse orthogonal transformation circuit 708, and then to the addition circuit. At 709, the same value as the value (predicted value) used when the prediction error is obtained by the subtraction circuit 703 is added and stored in the frame memory 711 or 712 via the selector switch 710 as locally decoded image data. Is done.

The frame memories 711 and 712 are frame memories having a function of storing the locally decoded image data in frame units as described above.

The locally decoded blocks stored in the frame memories 711 and 712 are stored in the control circuit 7 as described above.
Under the control of the control unit 16, the motion vector search circuit 717 via the selector switches 714 and 715 a when searching for a motion vector, and the selector switch 71 when calculating a prediction error.
4, 715b to the subtraction circuit 703, and similarly, at the time of local decoding, to the addition circuit 709 via the selector switches 714, 715c.

[0020]

As described above,
Since the conventional encoding apparatus does not have a function of using information accompanying the original image data, shape information (contour information) for identifying a background area and a target image area as represented by MPEG-4. Even when an image having information effective for motion compensation is encoded as described above, the above-mentioned information cannot be used, which has been a bottleneck in shortening the encoding processing time and improving the motion compensation accuracy.

Therefore, the present invention has been made in view of the above-mentioned problems of the conventional example. When an image having contour information is encoded, the encoding is performed by using this shape information. An object of the present invention is to improve the processing time and the motion compensation accuracy. Also, for images without contour information,
It is an object of the present invention to extract outline information from the image and use the outline information to perform encoding in a similar manner.

[0022]

In order to achieve the object of the present invention, an encoding apparatus according to the present invention is an encoding apparatus for encoding image data of each frame of a moving image. Storing means for storing a second frame that is temporally earlier than the first frame among the frames of the moving image; and reading out the second frame from the storing means. And a dividing unit for dividing the first frame and the first frame into blocks having a predetermined size, and contour information of objects included in the first frame and the second frame for each block by the dividing unit. Extracting means for extracting the outline information of the object included in the first frame and the outline information of the object included in the second frame, extracted by the extracting means. Then, a motion vector from the second frame to the first frame is calculated in block units, and the motion vector is used to calculate a motion vector temporally later than the first frame from the first frame. A prediction unit that predicts a motion vector to the third frame, which is a frame, in units of blocks; a block indicated by the motion vector predicted by the prediction unit in the third frame; Encoding means for encoding a difference value with respect to a block in the first frame corresponding to the first frame.

One of the encoding methods of the present invention is as follows.
An encoding method for encoding image data of each frame of a moving image, comprising:
Storing a second frame, which is a frame temporally earlier than the frame of the second frame, in a predetermined storage unit; reading the second frame from the predetermined storage unit; A dividing step of dividing one frame into blocks having a predetermined size; and an extracting step of extracting contour information of an object included in the first frame and the second frame for each block by the dividing step. From the frame 2 to the frame 1 using the outline information of the object included in the first frame and the outline information of the object included in the second frame extracted in the extraction step. A motion vector is calculated for each block, and the motion vector is used to calculate the frame 1 from the first frame.
A prediction step of predicting a motion vector to the third frame, which is a later frame, in units of blocks, and a block indicated by the motion vector predicted in the prediction step in the third frame. And an encoding step of encoding a difference value between the block and the block in the first frame that corresponds to the block in position.

A storage medium according to the present invention is a storage medium for storing a program code that functions as an encoding device that encodes image data of each frame of a moving image. A program code for a storage step of storing a second frame, which is a frame temporally earlier than the first frame, in a predetermined storage unit among the frames, and storing the second frame from the predetermined storage unit. A program code of a dividing step of reading and dividing the second frame and the first frame into blocks having a predetermined size; and a program code of the first frame and the second frame for each block by the dividing step. A program code of an extraction step for extracting contour information of an object included in the frame; and a program code included in the first frame extracted in the extraction step. Using the outline information of the object and the outline information of the object included in the second frame, a motion vector from the second frame to the first frame is calculated for each block, and the motion vector is calculated. A program code of a prediction step of predicting a motion vector from the first frame to the third frame, which is a frame temporally later than the first frame, on a block-by-block basis, 3
In the frame of the above, a program for an encoding step of encoding a difference value between a block indicated by a motion vector predicted in the prediction step and a block in the first frame corresponding to the block in position And a code.

Also, one of the encoding devices of the present invention is:
What is claimed is: 1. An encoding apparatus for encoding image data of each frame of a moving image, comprising: extracting means for extracting contour information of an object included in each frame; and extracting contour information of the object extracted by the extracting means. Detecting means for detecting a motion vector of image data of the encoding target frame using the encoding means, and encoding means for performing motion compensation encoding of the image data of the encoding target frame using the motion vector detected by the detecting means. It is characterized by having.

One of the encoding methods of the present invention is as follows.
An encoding method for encoding image data of each frame of a moving image, comprising: an extraction step of extracting outline information of an object included in each frame; and extracting the outline information of the object extracted by the extraction step. A detection step of detecting a motion vector of the image data of the encoding target frame using the motion vector, and an encoding step of performing motion compensation encoding of the image data of the encoding target frame using the motion vector detected by the detection step. It is characterized by having.

One storage medium of the present invention is a storage medium for storing a program code functioning as an encoding device for encoding image data of each frame of a moving image. A program code of an extraction step of extracting contour information of an included object; and a program code of a detection step of detecting a motion vector of image data of an encoding target frame using the contour information of the object extracted in the extraction step. A program code for an encoding step of performing motion compensation encoding on the image data of the encoding target frame using the motion vector detected in the detecting step is stored.

[0028]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail according to preferred embodiments with reference to the accompanying drawings.

[First Embodiment] FIG. 1 is a block diagram showing the configuration of an encoding apparatus according to this embodiment.

In FIG. 1, reference numeral 101 denotes a frame memory for an original image, 102 denotes a block circuit, 103 denotes a subtraction circuit,
104 is an orthogonal transformation circuit, 105 is a quantization circuit, 106 is a variable length coding circuit, 107 is an inverse quantization circuit, 108 is an inverse orthogonal transformation circuit, 109 is an addition circuit, 110, 114, and 1
15 is a selector switch, 111 and 112 are frame memories, 113 is an average value calculation circuit, 116 is a control circuit,
17 is a motion prediction circuit, 118 is a motion vector search circuit,
119 is an outline information buffer, and 120 is an edge detection circuit. The processing in the encoding device according to the present embodiment will be described with reference to the flowchart shown in FIG.

In FIG. 1, a frame memory 101 stores moving image data that has been edited and each frame includes moving image data composed of a plurality of independent layers and each frame of the original image data. Identification information (shape information) for identifying a background image area and an object image area is stored for each of the constituent layers.

First, under the control of the control circuit 116, original image data to be encoded is selected from the frame memory 101 and read. In parallel with the blocking process for the selected original image data in the blocking circuit 102, the control circuit 116 detects shape information corresponding to the layer in which each block is located from the blocking circuit 102.

When the shape information corresponding to the above-mentioned image of each layer constituting each frame read from the frame memory 101 is searched, and the shape information corresponding to the above-mentioned image of the layer is detected, Under the control of the control circuit 116, the shape information corresponding to the image of each layer constituting each frame is read out from the frame memory 101 and stored in the outline information buffer 119.

Under the control of the control circuit 116, the shape information is written in the outline information buffer 119 in a frame unit in parallel with the reading of the image of the layer corresponding to the shape information. The reading of the shape information from the frame memory 101 and the writing to the contour information buffer 119 are repeated every time a hierarchical image is encoded.

On the other hand, original image data selected under the control of the control circuit 116 and read out from the frame memory 101 is input to the blocking circuit 102 and divided into predetermined sizes (n × mn, m: constant). All subsequent processing is performed using the block as a unit. That is, the flowchart shown in FIG. 4 is a flowchart of the processing in block units.

The original image data (blocked image data) divided into blocks is input to the edge detection circuit 120. The processing of the input block image data in the edge detection circuit 120 differs depending on the presence or absence of shape information corresponding to the input block image data. In the present embodiment, the edge detection is shown in FIG. This is performed using the edge detection circuit 120 having the above structure.

Therefore, first, it is confirmed whether or not there is shape information corresponding to the input block image data (step S402).

If the shape information corresponding to the block image data input to the edge detection circuit 120 has been detected and stored in the contour information buffer 119, the edge detection processing in the edge detection circuit 120 Not done.

On the other hand, when the shape information corresponding to the block image data input to the edge detection circuit 120 is not detected, the edge detection circuit 120
Under the control of 6, the block image data is stored in the buffer 201.

Here, referring to FIG. 2, FIG.
According to the flowchart of the edge information detection process shown in FIG.
3) will be described.

The block image data stored in the buffer 201 is appropriately input to a DCT (discrete cosine transform) processing circuit 202 (step S501a), and is converted into DCT coefficients (step S502a).

The block image data converted into DCT coefficients is stored in the selected cell of the memory 203 in units of blocks under the control of the controller 206. The above processing is performed for all blocks (step S50).
3a, Step S504a).

D for each block stored in the memory 203
The CT coefficients are read under the control of the controller 206 and input to the edge detection unit 204. Edge detector 2
04 evaluates the frequency spectrum of each two-dimensionally continuous block image data (step S505a),
An edge extending over a plurality of blocks is estimated, and edge information is generated (step S506a). The generated edge information is input to the contour information generating circuit 205, and contour information (edge map) for each layer constituting each frame of the original image.
Is formed (step S404) and output (step S404).
405).

At this time, the edge map generated by the outline information generation circuit 205 is output to the outline information buffer 119 and the motion prediction circuit 117. The edge map input to the contour information buffer 119 is stored as a substitute for shape information that does not exist in the original image, and can be used when a current frame (hierarchical) image to be encoded is referred to by prediction. ing.

Edge detection circuit 120 having the above structure
Besides, the edge detection circuit 3 having the structure shown in FIG.
00 may be used.

First, if the shape information corresponding to the block image data input to the edge detection circuit 300 has been detected by the detection processing of the control circuit 116 and stored in the contour information buffer 119, the edge No edge detection processing is performed in the detection circuit 300.

On the other hand, when no shape information corresponding to the block image data input to the control circuit 116 is detected, the edge detection circuit 300 controls the block image data 300 under the control of the control circuit 116. The data is stored in the buffer 301.

FIG. 5B is a flowchart showing the processing in the edge detection circuit 300, and will be described.

The block image data stored in the buffer 301 is appropriately read out (step S501).
b), is input to the sum calculator 302, and calculates the sum of the luminance values of all the pixels constituting the input block image data (step S502b). The calculated sum of the luminance values of all pixels for each block is stored in the selected cell of the memory 303 in units of blocks under the control of the controller 306. The above processing is performed for all blocks (step S503b, step S504b).

The sum of the luminance values of all the pixels for each block stored in the memory 303 is read out under the control of the controller 306 and input to the edge detection unit 304. The edge detection unit 304 treats the sum of the luminance values of all pixels for each block as a parameter representing the DC component of the spatial frequency spectrum related to the image of this block, and compares and evaluates the DC component between two-dimensionally continuous blocks. (Step S505b), edge detection is performed, and edge information is generated (Step S506b).

The generated edge information is input to the contour information generating circuit 305, which forms and outputs contour information for each layer constituting each frame of the original image.

At this time, the outline information generated by the outline information generation circuit 305 is output to the outline information buffer 119 and the motion prediction circuit 117. Hereinafter, as in the case of the edge detection circuit 120 having the configuration shown in FIG. 2, the contour information input to the contour information buffer 119 is stored as a substitute for shape information that does not exist in the original image, and Is available when the frame (hierarchical) image is predicted and referenced.

Next, the control circuit 116 reads out the contour information stored in the contour information buffer 119 (step S406) and inputs it to the motion prediction circuit 117. The operation of the motion prediction circuit 117 will be described with reference to FIG.

As described above, the contour information is the shape information itself if the shape information corresponding to the original image exists, and if not, the contour information is generated by the edge detection circuit 120 as described above. Edge map (Fig. 6
(A)). Subsequently, the edge map (FIG. 6B) corresponding to the image to be motion predicted is stored in the contour information buffer 1.
19 (step S407), and a comparative evaluation is performed for each frame (or layer), and the frame (or layer) is evaluated.
The predicted value MVx of the motion vector in block units of FIG.
(See (a)) is calculated (step S408).

The calculated predicted value MVx of the motion vector
Is input to the motion vector search circuit 118. The motion vector search circuit 118 calculates a motion vector search start point (FIG. 8B) from the input motion vector predicted value MVx.
Then, a motion vector search is performed in the arbitrarily set motion vector search range in the same manner as in the conventional example around the calculated motion vector search start point (step S409). The motion vector search start point in the present embodiment is defined as a position offset from the current position of the current block to be encoded by the horizontal and vertical coordinates corresponding to the predicted value MVx of the motion vector.

After the motion vector of the block to be coded is determined as described above, the process of coding the block to be coded using this motion vector (step S
410) will be described with reference to the flowchart shown in FIG.

After the motion vector of the encoding target block is determined as described above, the encoding target block is input from the blocking circuit 102 to the subtraction circuit 103 under the control of the control circuit 116. At the same time, a block (prediction reference block) corresponding to the position specified by the motion vector is determined from the reference frame (locally decoded) corresponding to the encoding target block (step S901), and this prediction reference block is stored in the frame memory 111. Or 112, and inputs the result to the subtraction circuit 103,
In step S90, subtraction is performed between the corresponding pixels of the prediction reference block from the encoding target block (step S90).
2).

Subsequent processing is the same as in the conventional example, in which the block to be coded, which is a difference value (prediction error) block calculated by the subtraction circuit 103, is converted into an orthogonal transform coefficient block by the orthogonal transformer 104. (Step S903).
The encoding target block, which has been subjected to the coordinate transformation into the orthogonal transformation coefficient by the orthogonal transformation circuit 104, is input to the quantization circuit 105, and each coefficient of the encoding target block is quantized with a predetermined quantization coefficient (step S904). . The quantization coefficient can be arbitrarily selected and controlled according to conditions such as a desired code amount and image quality.

The quantized block to be coded is input to the variable-length coding circuit 106, and a variable-length code is added to each coefficient of the block to be coded after the quantization process, and the resulting block is output. (Step S905).

The block to be coded which has been quantized at the same time is input to the inverse quantization circuit 107, and the coefficients used in the block to be coded which have been quantized are used by the quantization circuit 105. An inverse quantization process is performed by multiplying the coefficients. The block to be coded, which has been inversely quantized by the inverse quantization circuit 107, is input to the inverse orthogonal transform circuit 108, which performs inverse orthogonal transform paired with the orthogonal transform process in the orthogonal transform circuit 104, and performs local decoding of the prediction error A local decoding prediction error block, which is a completed block, is generated and input to the adding circuit 109. The local decoding prediction error block input to the addition circuit 109 uses the prediction reference block used for obtaining the prediction error in the subtraction circuit 103 under the control of the control circuit 116 to control the selector switches 114 and 1.
The signal is read out again via the line 15c, and is added to the corresponding pixel of the local decoding prediction error block.

The local decoding prediction error block added to the prediction reference block is stored and held in the frame memory 111 or 112 via the selector switch 110 as a local decoding block. The frame memory 111,
Reference numeral 112 denotes a frame memory having a function of storing locally decoded image data in frame units.

The local decoding blocks stored in the frame memories 111 and 112 are stored in the control circuit 11 as described above.
6, the motion vector search circuit 1 is connected via the selector switches 114 and 115a during the motion vector search.
17, when the prediction error is calculated, the selector switch 114,
The signals are output to the subtraction circuit 103 via 115b.

[Other Embodiments] Even if the above-described embodiment is applied to a system composed of a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), it is composed of one device. The present invention may be applied to an apparatus (for example, a copying machine, a facsimile machine, etc.).

Further, an object of the above-described embodiment is to supply a storage medium (or a recording medium) in which program codes of software for realizing the functions of the above-described embodiment are recorded to a system or an apparatus, and to provide the system or apparatus with the storage medium. It is needless to say that the present invention is also achieved when a computer (or a CPU or an MPU) reads and executes a program code stored in a storage medium. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiment, and the storage medium storing the program code constitutes the above-described embodiment. By executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also an operating system (OS) running on the computer based on the instruction of the program code. It goes without saying that a case where some or all of the actual processing is performed and the functions of the above-described embodiments are realized by the processing is also included.

Further, after the program code read from the storage medium is written into the memory provided in the function expansion card inserted into the computer or the function expansion unit connected to the computer, the program code is read based on the instruction of the program code. Needless to say, the CPU included in the function expansion card or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

When the above embodiment is applied to the storage medium, the storage medium stores program codes corresponding to the above-described flowcharts (shown in FIGS. 4 and 5).

[0067]

As described above, according to the present invention,
By performing encoding using this shape information when encoding an image having contour information, encoding processing time and motion compensation accuracy can be improved. Also, for an image having no outline information, encoding can be performed in the same manner by extracting outline information from this image and using this outline information.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an encoding device according to a first embodiment of the present invention.

FIG. 2 is a diagram illustrating a structure of an edge detection circuit.

FIG. 3 is a diagram illustrating a structure of an edge detection circuit.

FIG. 4 is a flowchart of a process in the encoding device according to the first embodiment of the present invention.

FIG. 5 is a flowchart of a process in an edge detection circuit.

FIG. 6 is a diagram illustrating the operation of the motion prediction circuit.

FIG. 7 is a block diagram illustrating an example of a conventional encoding device.

FIG. 8 is a diagram showing a motion vector search method.

FIG. 9 is a flowchart of a process of encoding a current block.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C059 KK11 MA05 MA23 MB00 MB03 ME01 NN01 NN03 NN21 NN28 NN38 SS20 SS26 SS28 TA61 TC01 TC10 TC34 UA33 5J064 AA02 BA01 BA09 BA16 BB03 BC01 BC16 BC21 BD01 5L096 AA08 BA07 FA06 GA06

Claims (15)

[Claims] 1. An encoding apparatus for encoding image data of each frame of a moving image, wherein the encoding device is a frame which is a temporally earlier frame than a first frame among the frames of the moving image. Storage means for storing the second frame, a division means for reading the second frame from the storage means, and dividing the second frame and the first frame into blocks having a predetermined size; For each block by the dividing means, the first frame,
Extracting means for extracting outline information of an object included in the second frame; outline information of an object included in the first frame extracted by the extracting means; and an object included in the second frame A motion vector from the second frame to the first frame is calculated in block units using the contour information of the first frame, and the first frame is calculated from the first frame using the motion vector. A prediction unit that predicts a motion vector to the third frame, which is a later frame, in units of blocks; and a block indicated by the motion vector predicted by the prediction unit in the third frame. A coding method for coding a difference value between a block in the first frame and a block corresponding to the block in the first frame. Encoding apparatus characterized in that it comprises and. 2. The method according to claim 1, wherein the extracting unit further determines whether or not the contour information of the object exists in the first frame and the second frame. When there is no object contour information in the second frame,
2. An edge detection unit according to claim 1, further comprising: an edge detection unit that performs edge detection on the first frame and the second frame, and uses the edge detection result as the contour information of each frame. Encoding device. 3. The edge detecting means performs a DCT transform on the first frame and the second frame,
The encoding apparatus according to claim 2, wherein the edge information of each frame is generated using edge information generated by evaluating each DCT coefficient. 4. The edge detecting means obtains a sum of luminance values of the first frame, the second frame, and each of the brightness values, and uses edge information generated by using the sum,
3. The encoding apparatus according to claim 2, wherein contour information of each frame is generated. 5. The predicting unit further includes: a search start point based on a motion vector of a block of interest from the second frame to the first frame, and an area having a predetermined size, Area setting means for setting a frame of the target block; a block specifying means for specifying, as a prediction block for the target block, a region most similar to the target block in the region set by the region setting means; and specifying the block from the target block. The encoding apparatus according to any one of claims 1 to 4, further comprising: a calculating unit configured to calculate a motion vector for the prediction block specified by the unit. 6. The encoding apparatus according to claim 1, wherein each frame has a hierarchical structure, and the outline information can be stored for each hierarchical level. 7. An encoding method for encoding image data of each frame of a moving image, wherein the first frame is a temporally earlier frame than the first frame among the respective frames of the moving image. Storing the second frame in a predetermined storage unit, reading the second frame from the predetermined storage unit, and converting the second frame and the first frame into blocks having a predetermined size. A dividing step of dividing, the first frame for each block by the dividing step,
An extraction step of extracting outline information of an object included in the second frame; an outline information of an object included in the first frame, extracted in the extraction step; and an object included in the second frame. And a motion vector from the second frame to the first frame is calculated in block units using the contour information of the first frame and the first frame is calculated from the first frame using the motion vector. A prediction step of predicting a motion vector to the third frame, which is a frame temporally later than the frame, in units of blocks; and a block indicated by the motion vector predicted in the prediction step in the third frame. And encoding for encoding a difference value between a block in the first frame and a block corresponding to the block in position. Coding method characterized by and a degree. 8. The extracting step further includes: a determining step of determining whether or not contour information of an object exists in the first frame and the second frame; and determining the first frame by the determining step. When there is no object contour information in the second frame,
8. An edge detection step of performing edge detection on the first frame and the second frame, and using the edge detection result as the contour information of each frame. Encoding method. 9. The predicting step further includes: setting a region having a predetermined size around a search start point based on a motion vector of a block of interest from the second frame to the first frame; An area setting step of setting a frame of the target block; and a block specifying step of specifying, as a prediction block for the target block, a region most similar to the target block in the region set by the target area setting step; The encoding method according to claim 7, further comprising: a calculating step of calculating a motion vector for the predicted block specified in the step. 10. A storage medium for storing a program code that functions as an encoding device that encodes image data of each frame of a moving image, wherein: A program code for a storage step of storing a second frame, which is a temporally previous frame, in a predetermined storage means; and reading the second frame from the predetermined storage means, A program code for a dividing step of dividing one frame into blocks having a predetermined size; and a first frame for each block by the dividing step.
A program code of an extraction step for extracting outline information of an object included in the second frame; an outline information of an object included in the first frame, extracted in the extraction step; The motion vector from the second frame to the first frame is calculated in block units using the contour information of the included object, and the motion vector is used to calculate the motion vector from the first frame. A program code of a prediction step of predicting a motion vector to the third frame, which is a frame temporally later than the first frame, in block units, and a prediction of the third frame in the prediction step A block indicated by a motion vector and a block in the first frame corresponding to the block in position Storage medium characterized by comprising a program code for a coding step for performing coding on the difference value. 11. The program code of the extracting step further includes: a program code of a determining step of determining whether or not contour information of an object exists in the first frame and the second frame; , When there is no object contour information in the first frame and the second frame,
11. A program code for an edge detection step of performing edge detection on the first frame and the second frame, and using the edge detection result as the contour information of each frame. A storage medium according to claim 1. 12. The program code of the prediction step further includes: setting an area having a predetermined size around a search start point based on a motion vector of a target block from the second frame to the first frame; A program code of an area setting step to be set in the third frame; and a program code of a block specifying step of specifying an area most similar to the target block as a prediction block for the target block in the area in the area setting step The storage medium according to claim 10, further comprising: a program code for calculating a motion vector from the block of interest to the prediction block specified in the block specifying step. 13. An encoding device for encoding image data of each frame of a moving image, comprising: extraction means for extracting outline information of an object included in each frame; Detecting means for detecting the motion vector of the image data of the encoding target frame using the contour information of the object; and performing motion compensation encoding on the image data of the encoding target frame using the motion vector detected by the detecting means. An encoding device comprising: an encoding unit. 14. An encoding method for encoding image data of each frame of a moving image, comprising: an extraction step of extracting outline information of an object included in each frame; and an extraction method extracted by the extraction step. A detection step of detecting a motion vector of the image data of the encoding target frame using the contour information of the object; and a motion compensation encoding of the image data of the encoding target frame using the motion vector detected in the detection step. An encoding step. 15. A storage medium for storing a program code that functions as an encoding device that performs encoding on image data of each frame of a moving image, the extraction medium extracting contour information of an object included in each frame. A program code of a step, a program code of a detecting step of detecting a motion vector of image data of the encoding target frame using the contour information of the object extracted in the extracting step, and a motion vector detected by the detecting step. And a program code for an encoding step for performing motion compensation encoding on the image data of the encoding target frame using the encoding method.