JP2001251628A - Coder for dissolve picture - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a coder that conducts no motion retrieval processing or conducts moving picture compression coding processing at a high-speed for a motion retrieval processing with a small arithmetic amount in the case of receiving a compression coded moving picture and generating a compression coded moving picture whose dissolve is realized for a transition period. SOLUTION: Motion compensation information decided by two input pictures is used for motion compensation information for re-coding of a synthesis picture for a dissolve period. Motion compensation information used for a received picture with a high input ratio and motion compensation information for a received picture with a value highly reflecting a feature quantity of the received picture in addition to the input ratio are used for determination of the motion compensation information. Furthermore, a coding system identical to that for a picture to reference motion compensation information is adopted for each synthesized picture within the dissolve period and timing to switch the picture to reference the motion compensation information is shifted depending on number of adjacent B frames to suppress deterioration in the coding efficiency and to reduce the entire arithmetic quantity.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dissolve picture coding apparatus, and more particularly to a dissolve picture coding apparatus using inter-frame motion compensation predictive coding.

[0002]

2. Description of the Related Art As one of special effects inserted at the time of a scene change of a moving image, a time transition section is provided in the scene change, and as shown in FIG. There is a “dissolve” that changes from a previous image (moving image A) to an image after a scene change (moving image B). In the dissolve, the image in the transition section is a composite image of two input images.

[0003] A digital moving image is represented by a sequence of continuous digital still images. In addition, a digital still image is represented by a set of pixels arranged at equal intervals in space. When “dissolve” is used in a digital moving image, each pixel value of the composite image is represented by the following equation (1) (see FIG. 2).

Dt (y, x) = {(tout−t) / (tout−tin)} · At (y, x) + {(t−tin) / (tout−tin)} · Bt (y, x …… (1)

Here, Dt (y, x) is the pixel value of the coordinates (y, x) in the composite image at the time t, and At (y, x) is the coordinates (y, y) in the image before the switching of the time t. , X), B
t (y, x) is the pixel value of the coordinates (y, x) in the image after the switching of the time t, tin is the transition section start time, and tout
Is the transition section end time.

[0006] However, for moving image information, compression coding of the image information is actively used due to the large amount of information. In particular, in order to increase the compression efficiency, inter-frame prediction coding using motion compensation prediction information from information of temporally neighboring images is used. MPEG video coding and the like use this method.

[0007] In order to create a composite image using a dissolve effect in a transition section from a moving image that has been subjected to compression encoding using such inter-frame prediction encoding, for example, it is shown in FIG. It is necessary to make such operation. First, a coded moving image A61 as a target is decoded and reproduced by a first moving image decoding unit 63, and a coded moving image B62 is decoded.
Is decoded and reproduced in the second video decoding unit 64. The decoded and reproduced image information A and B are sent to the image synthesizing unit 65. In the image synthesizing section 65, a synthetic image creating process for the dissolve section is performed, and the created synthetic image D is sent to the moving image re-encoding section 66. The moving image re-encoding unit 66 re-encodes the moving image from the input synthesized image D using inter-frame prediction, and outputs an encoded moving image E67.

FIG. 18 shows a moving picture re-encoding unit 6 shown in FIG.
FIG. 6 is a block diagram illustrating the details of a sixth example. From the input composite image D, the motion search unit 71 performs a motion search process with the already processed image stored in the image storage unit 72, and performs motion compensation used when re-encoding the composite image. The information a is determined. Next, using the obtained motion compensation information a and the image b stored in the image storage unit 72, a motion compensation image c is created in the motion compensation unit 73.

Intra-frame / inter-frame coding determination section 74
Then, whether to perform intra-frame encoding or inter-frame encoding is determined based on the features of the motion compensation image c and the input composite image D. According to the determination result, the changeover switch S
, And in the case of intra-frame encoding, the input synthesized image D
In the case of inter-frame encoding, an image obtained by subtracting the motion compensation image c from the input composite image D is output as motion compensation prediction error information d. The motion compensation prediction error information d
Is converted using a discrete cosine transform or the like in the conversion unit 75 to become conversion region information e. The conversion region information e is coded in the coding unit 76 and output as a coded moving image E.

On the other hand, the conversion area information e output from the conversion unit 75 is returned to the original motion compensation prediction error amount by the inverse conversion unit 77, and is added to the motion compensation image to become a decoded reproduced image f, and the image is stored. The image is stored in the unit 72 and is used as a reference image in the subsequent image encoding.

[0011]

However, in the above-mentioned conventional method, the moving picture re-encoding unit 66 after dissolve image synthesis includes a motion search unit 71. The motion search process performed by the motion search unit 71 is a process of searching for an image of a target image or a small block in the target image that is similar to an image of which part in a reference image. Requires computational complexity. Therefore, the conventional method has a problem that it takes a long time to re-encode the composite image. In addition, in a configuration without the motion search unit 71, there is a problem that a moving image cannot be re-encoded.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems of the prior art, and to perform a motion search process when creating a compressed coded moving image in which a dissolution is realized in a transition section by using the compressed coded moving image as an input. An object of the present invention is to provide a dissolve image encoding apparatus capable of performing a moving image compression encoding process at a high speed without performing the operation or greatly reducing the operation amount of the motion search process.

[0013]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention provides a method for decoding motion compensation information at the time of re-encoding a synthesized image without performing a motion search process. A point that a unit for extracting motion compensation information from a moving image and determining motion compensation information in the synthesized image from the extracted motion compensation information and temporal position information of the synthesized image to be processed is provided. There is one feature.

A second feature of the present invention is that the present invention includes means for improving coding efficiency by performing a small area motion search around a position indicated by the determined motion compensation information as a starting point. is there.

A third feature of the present invention resides in that a means for using motion compensation information of an image having a high input ratio of two input images in each composite image is provided.

Further, according to the present invention, an input image which uses the motion compensation information based on the result of reflecting the characteristic amount of each input image on an input ratio serving as a reference for selecting motion compensation information of two input images. There is a fourth feature in that a means for determining is provided.

Further, the present invention is characterized in that a means for using pattern information is provided as a feature amount of an input image, and a means for using a variance in an image or an absolute error sum with an average value in an image is provided. There is a fifth feature.

A sixth feature of the present invention resides in that, in the vicinity of the center of the dissolve section, when the difference between the input ratios of the two input images is small, there is provided means for selecting an intra-picture encoding method. is there.

The seventh feature of the present invention resides in that the present invention further comprises means for making the same as the encoding method of the image referring to the motion compensation information when re-encoding the synthesized image.

An eighth feature is that a means for changing the switching time of the motion-compensation-information reference image obtained by the calculation according to the encoding method of the neighboring frame is provided.

Further, according to the present invention, when an encoded dissolve image is created, a method of decoding input encoded image data up to pixel region information and re-encoding after a synthesizing operation, and a method of transforming the input image data into transform region information Ninth point in that a means for selecting a method of decoding and re-encoding after the combining operation is provided.
There is a feature.

Further, according to the present invention, when a prediction frame interval differs between an image referencing motion compensation information and an image to be re-encoded, the motion compensation information is scaled according to the ratio of the prediction frame intervals. A tenth feature resides in that means are provided.

According to the above-mentioned feature, the re-encoding process can be performed without using the motion search process which requires a very large amount of calculation which has been performed by the conventional method, thereby solving the conventional problem. be able to.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the drawings. FIG. 3 is a block diagram showing an embodiment of the encoding device of the present invention.

The encoded moving image information before being combined by the dissolve effect is referred to as an encoded moving image A and an encoded moving image B. The first and second moving image decoding units 1 and 2 decode the input coded moving images A and B, respectively. The video decoding units 1 and 2 respectively extract motion compensation information p and q such as motion vector information and predicted frame interval information from the coded image information sequentially inputted, and
Send to The motion compensation information p, q indicates which part of the reference image refers to which part of the target image or the entire spatial part.

In the motion compensation information decision unit 3, the motion compensation information p, q extracted from the two encoded moving pictures A, B
, The motion compensation information r in the dissolve section is determined. Further, the determined motion compensation information r is sent to the moving picture re-encoding unit 5. On the other hand, the image combining unit 4 combines the decoded and reproduced images A1 and B1 in the dissolve section to generate a combined image D1.

The moving picture re-encoding unit 5 performs motion compensation and coding of the input composite image D 1 using the motion compensation information r input from the motion compensation information determining unit 3, and generates a generated code. And outputting the coded moving image E.

FIG. 4 is a block diagram for explaining details of the moving picture decoding unit 1 of FIG. The input encoded image A is decoded by the decoding unit 11. Of the decoded information, the transform area information p1 is sent to the inverse transform unit 12, where the inverse transform unit 1
In 2, an orthogonal transform such as an inverse discrete cosine transform is performed.
The motion compensation prediction error information p2 is reproduced. On the other hand, the motion compensation information p decoded by the decoding unit 11 is sent to the motion compensation unit 14 and the motion compensation information determination unit 3.

The motion compensating unit 14 performs a motion compensating process from the motion compensating information p and the already decoded image data stored in the image storing unit 13, and obtains the image p obtained by the process.
An image is reproduced by the sum of 3 and the motion compensation prediction error information p2, and is output as a decoded reproduced image A1. The moving image decoding unit 2 in FIG. 3 also performs the same processing as the moving image decoding unit 1 on the encoded moving image B, and outputs a decoded reproduced image B1.

FIG. 5 is a block diagram illustrating the details of the moving picture re-encoding unit 5 of FIG. The moving picture re-encoding unit 5
Is composed of an intra-frame / inter-frame coding determination unit 21, a conversion unit 22, a coding unit 23, an inverse conversion unit 24, an image storage unit 25, and a motion compensation unit 26. The presence / absence of the motion search unit 71 differs from that of the encoding unit 66. In the conventional device, the motion compensation information is obtained by a motion search process between the image stored in the image storage unit 72 and the composite image D. On the other hand, in the embodiment of the present invention, the motion compensation processing is not performed as shown in FIG. 5 and the motion compensation unit 26 uses the motion compensation information r input from the motion compensation information determination unit 3 to perform the motion search. Perform compensation processing.

FIG. 6 is a diagram illustrating a specific example of a method of determining motion compensation information in a dissolve section performed by the motion compensation information determining unit 3. Hereinafter, the encoding process of the dissolve image will be described.

A scene is switched using a dissolve in a transition area of an image that changes from the input coded video A to the input coded video B, and a new coded video E is created.
Assuming that the time tin is the dissolve start time and the time tout is the dissolve end time, a period between tin and tout is a dissolve section.

The frame numbers 1 to 1 in the encoded moving image E
No. 3 is the image before the dissolve section, and is the same as the information of the input encoded moving image A. Frame numbers 4 to 7 are frames existing in the dissolve section. In the figure, the input ratio (%) of the input coded video A and the input coded video B in each frame is shown. The input coded moving image A and the input coded moving image B are respectively decoded, and a combined image D1 is created according to the input ratio of each frame. After that, the composite image is re-encoded to create an encoded moving image E.

Of these frames, in frame numbers 4 and 5, the input ratio of the input coded video A is higher than the input ratio of the input coded video B. Therefore, when re-encoding these composite images, the motion compensation information of the input encoded moving image A is used. On the other hand, in frame numbers 6 and 7, the input ratio of the input coded video B is higher than the input ratio of the input coded video A. Therefore, the motion compensation information of the input encoded moving image B is used when re-encoding these composite images. Frame numbers 8 to 10 in the composite image D1 are:
Since the image is after the dissolve section, the information is the same as the information of the input encoded moving image B.

In FIG. 6, the input coded image utilizing the motion compensation information is switched between frame numbers 5 and 6. Assuming that the switching time is tc, tc can be expressed by the following equation (2-1). Further, the motion compensation information r (= MV) E of the synthesized coded video E is expressed by the following equation: (2-2)
Can be represented by Here, MVA and MVB represent the motion compensation information of the coded input video A and the coded input video B, respectively. tc = tin + (tout−tin) / 2 (2-1) if t ≦ tc, MVE = MVA, else MVE = MVB (2-2)

It is also possible to set tc in the above equation not according to the input ratio but according to the feature amount of two input images. When determining the time at which the reference image is switched according to the input image feature, it can be obtained by the following equation (3). Here, FA and FB are, respectively,
The feature amounts of the coded input image A and the coded input image B in the dissolve section are shown. tc = tin + FA (tout-tin) / (FA + FB) (3)

As the feature quantity in the above equation, a value reflecting the pattern information of each input coded image can be used. As a value reflecting the pattern information, a pixel variance value in a block for each coded block of each coded input image, an absolute error sum with an average value in the block, and an absolute value of a coefficient in transform domain data such as DCT It is possible to reflect a value such as a sum.

FIG. 7 shows a second embodiment of the present invention.
7, the same reference numerals as those in FIG. 3 denote the same or equivalent components.

In this embodiment, the motion compensation information determining unit 3
Is characterized in that not only the motion compensation information p and q of the input image but also the image feature information s of the input image are used to determine the motion compensation information r. In the present embodiment, the decoded and reproduced images A1 and B1 output from the video decoding unit 1 and the video decoding unit 2 are also transmitted to the image feature analysis unit 7, and the image feature analysis unit 7 The feature amount is calculated, for example, the pattern information is calculated, and the image feature information s as the calculation result is transmitted to the motion compensation information determination unit 3.

Furthermore, instead of switching the coded input image using the motion compensation information at only one point in the dissolve section, coding using the motion compensation information by the product of the input ratio and the characteristic amount as shown in the following equation: A method of selecting an input image is also possible. if FA × RA> FB × RB, MVD
= MVA, else MVD = MVB where RA,
RB represents the input ratio of the input image A and the input image B at the same time in the dissolve section.

Next, the encoding method of each image in the dissolve section will be described. In a moving picture coding method for performing inter-frame predictive coding, the following three kinds of image coding methods are used in combination from the viewpoint of high efficiency coding and convenience of decoding and reproduction.

P frame: unidirectional prediction frame. One of the inter prediction frames. The image is encoded by inter-frame motion compensation prediction encoding from an image encoded in the past.
This prediction is forward prediction. The image is decoded and played back,
It becomes a reference image for the next P frame encoding. If the similarity between the referenced image and the referenced image is high, the coding efficiency is improved.

B frame: bidirectional prediction frame. One of the inter-frame prediction pictures. It is encoded by inter-frame motion compensation prediction from two temporally preceding and succeeding encoded images. The frame is not used as a reference image.

I frame: an intra-coded frame.
One image is independently encoded without using the motion and correlation between images. Therefore, independent frame decoding is possible.

The present invention is characterized in that the coding method of each frame in the dissolve section is the same as the coding method of an image that refers to the motion compensation information. This makes it possible to reduce the deterioration of the image quality at the time of re-encoding. Furthermore, the timing tc for switching the image referring to the motion compensation information in the dissolve section is changed to a position t'c where the total number of consecutive B frames adjacent before and after tc decreases. This makes it possible to reduce the number of times of scaling processing of motion compensation information, which will be described later, and reduce the deterioration of image quality and the number of calculations.

FIG. 8 shows the above explanatory diagram. FIG. 9A shows a case where tc is the center of 10 frames included in the dissolve section. Therefore, the first five frames use the coding scheme and motion compensation information of each frame of the input coded video A, and the last five frames use the coding scheme and motion compensation information of each frame of the input coded video B. I do. In this case, the motion compensation information in the five images surrounded by the line x needs to perform scaling of the motion compensation information described later because the prediction frame interval changes.

On the other hand, FIG. 2B shows the position of tc, and FIG.
This is an example in which it is delayed to t'c one frame behind from. In this example, the number of frames that require scaling of the motion compensation information is two, so that it is possible to reduce the deterioration of the image quality and improve the processing speed as compared with the example.

Next, scaling of motion vector information, which is processing when the predicted frame interval between the input image and the re-encoded image is different, will be described. The motion compensation information is represented by the motion vector information and the predicted frame interval information. As described above, the predicted frame interval in the input encoded image may be different from the predicted frame interval in the re-encoding of the dissolve composite image.

In such a case, the scaling of the motion vector information is performed according to the ratio of the prediction frame intervals. FIG. 9 illustrates the scaling of a motion vector according to the ratio of predicted frame intervals. FIG. 7A shows that the predicted frame interval is 3
This shows a case where encoding is performed at a predicted frame interval of two frames from an input image of a frame. In this case, 2/3 scaling is applied to the motion vector information.

When the prediction directions are different from each other, such as forward prediction and backward prediction, it is possible to cope by inverting the vector by multiplying by a negative value. FIG. 9 (b) of FIG. 9 shows a case in which motion vector information having a predicted frame interval of 1 frame is generated from a motion vector having a frame interval of 2 frames in backward prediction. Therefore, scaling of -1/2 is performed on the motion vector information.

Further, in the encoding of each image block,
In addition to the motion compensation prediction coding that performs motion compensation prediction between frames, there is an intra-frame coding method that performs coding without performing motion compensation prediction. In this case, the target area has no motion compensation information. In the present invention, when no motion compensation information exists at the target position of the referred coded input image, the motion vector information is set to a zero vector, and this is used as the motion compensation information.

Next, a third embodiment of the present invention will be described with reference to FIG. 10, the same reference numerals as those in FIG. 3 indicate the same or equivalent components.

The encoded moving image A is transmitted to the first moving image
1 and output as decoded conversion area information A2. The decoded conversion area information A2 is conversion area information obtained by using motion compensation in the conversion area from the motion compensation information and the conversion area information of the motion compensation prediction error. For example, in the MPEG coding method, it is represented by a DCT coefficient. Also, the encoded moving image B is decoded by the second moving image decoding unit 32, and is output as decoded conversion area information B2. Further, the motion compensation information p, q is sent to the motion compensation information determination unit 3.

Next, based on the decoded conversion area information A2 and the decoded conversion area information B2 output from the moving picture decoding section 31 and the moving picture decoding section 32, the conversion area image synthesizing section 33 outputs a conversion area which is a synthesized image of a dissolve section. Create composite image information D2. The moving image re-encoding unit 34 re-encodes the transformed area combined image information D2 using the motion compensation information sent from the motion compensation information determining unit 3. Then, the re-encoded information is output as encoded moving image information E.

FIG. 11 shows the moving picture decoding unit 3 in FIG.
FIG. 2 is a block diagram illustrating details of the first embodiment. Encoded video A
Is first decoded by the decoding unit 41, and the motion compensation information p is sent to the motion compensation information determination unit 3 and the transform area motion compensation unit 42. Further, the decoding unit 41 outputs the motion compensation prediction error conversion area information u.

Using the motion compensation information p, the conversion area motion compensation unit 42 performs motion compensation processing in the conversion area on the conversion area information of the already decoded image information stored in the conversion area information storage unit 43. . The conversion area motion compensation information v obtained by the motion compensation processing is added to the motion compensation prediction error conversion area information u output from the decoding unit 41, and output as decoded conversion area information A2. The second video decoding unit 32 in FIG. 10 performs the same process on the coded image B as the first video decoding unit 31, and outputs decoded conversion area information B2.

FIG. 12 shows the moving picture re-encoding unit 34 of FIG.
FIG. The motion compensation information r sent from the motion compensation information determination unit 3 is
1 in the transform area motion compensator 51.
The decoded area conversion image information already processed and stored in the conversion area image information storage unit 52 is subjected to motion compensation on the conversion area using the motion compensation information r, and the compensation result is converted to the conversion area motion compensated image information y. To the intra-frame / inter-frame coding determination unit 53.

Intra / interframe coding determination section 53
Predicts whether the coding efficiency of intra-frame coding or inter-frame coding will be higher based on the input transform region motion compensation image information y and the image transform region information, and compares the result with the intra-frame / It is sent to the inter-frame coding switching unit S. The switching unit S encodes the image conversion area information when performing intra-frame encoding, and the difference w between the image conversion area information and the conversion area motion compensation image information y when performing inter-frame encoding. Output to 54. In the encoding unit 54,
The variable length encoding process is performed, and the result is output as encoded moving image information E.

Further, the motion compensation prediction error conversion area information w input to the encoding unit 54 is added to the conversion area motion compensation information y output from the conversion area motion compensation unit 51, and the resultant is obtained as the decoded conversion area image information z. Conversion area image information storage unit 52
Is accumulated in In the present embodiment, since there is no process for performing the orthogonal transform and the inverse orthogonal transform, the amount of calculation can be significantly reduced, and high-speed processing can be performed.

FIG. 13 is a block diagram showing a fourth embodiment of the present invention. In this embodiment, the coded input image A is used for the image near the temporal center in the dissolve section.
The difference between the input ratios of the input image and the coded input image B is small, and the correlation is low with either of the input coded images. It is characterized in that the intra-frame coding method is used.

As shown in the figure, the motion compensation information determining section 3 outputs an intra-frame coding control signal r 1 to the moving picture re-encoding section 5 for an image near the temporal center in the dissolve section, For images other than near the temporal center in the dissolve section, the motion compensation information determined in the first to third embodiments is used. 3 and 7 in FIG.
And the same reference numerals as those in FIG. 10 indicate the same or equivalent components.

FIG. 14 is a block diagram showing details of the moving picture re-encoding unit 5 of FIG. At the time of re-encoding of the composite image D1, the intra-frame / inter-frame encoding determination unit 21 performs a process of determining whether to perform intra-frame encoding or inter-frame encoding. When the intra-frame coding control signal r1 is sent, the intra-frame coding system is forcibly selected, and the changeover switch S is switched to the intra-frame coding system.

Next, a fifth embodiment of the present invention is shown in FIG. In this embodiment, the motion compensation information r sent from the motion compensation information determination unit 3 is input to the small area motion search unit 27. Now, the motion compensation information sent from the motion compensation information determining unit 3 is the motion compensation information x shown in FIG.
m, the small area motion search section 27 performs a motion search process on the small area x'm centered on the motion compensation information xm, and finds the optimum position in the small area x'm with the new motion compensation information r2.
To the motion compensator 26. According to this embodiment, while the search area of the conventional motion search unit 71 in FIG. 18 is, for example, xs, it may be a small area x′m,
The amount of computation required for motion search can be significantly reduced without lowering the accuracy of motion compensation.

Note that all the processing of the first to fourth embodiments, for example, the processing of the motion compensation information determining unit 3 and the like can be applied to the fifth embodiment, but the description thereof is omitted.

[0065]

As is apparent from the above description, according to the present invention, when performing a process of creating an encoded moving image in which a dissolve effect is inserted from two pieces of encoded moving image information, a moving image re-encoding is performed. The above processing can be performed even if the conversion unit does not have the motion search unit. Further, even if a motion search is performed, a high-speed process can be realized because a motion search process with a small amount of arithmetic processing may be performed.

[Brief description of the drawings]

FIG. 1 is a diagram showing an outline of a dissolve.

FIG. 2 is a diagram illustrating a transition state in a dissolve section.

FIG. 3 is a block diagram showing a first embodiment of the dissolve image encoding device of the present invention.

FIG. 4 is a block diagram illustrating details of a moving image decoding unit in FIG. 3;

FIG. 5 is a block diagram illustrating details of a moving picture re-encoding unit in FIG. 3;

FIG. 6 is an explanatory diagram showing a method for determining motion compensation information according to the present invention.

FIG. 7 is a block diagram showing a configuration of a second embodiment of the present invention.

FIG. 8 is a diagram illustrating a change of a motion compensation information reference image switching time according to the present invention.

FIG. 9 is a diagram illustrating scaling of motion compensation information according to the present invention.

FIG. 10 is a block diagram showing a configuration of a third embodiment of the present invention.

11 is a block diagram illustrating details of a moving image decoding unit in FIG. 10;

FIG. 12 is a block diagram illustrating details of a moving picture re-encoding unit in FIG. 10;

FIG. 13 is a block diagram showing a configuration of a fourth embodiment of the present invention.

14 is a block diagram illustrating details of a moving picture re-encoding unit in FIG. 13;

FIG. 15 is a block diagram illustrating a configuration of a video transcoder according to a fifth embodiment of the present invention.

FIG. 16 is a diagram illustrating a small area motion search method according to the present invention.

FIG. 17 is a block diagram showing a configuration of a conventional device.

FIG. 18 is a block diagram illustrating details of a re-encoding unit of the conventional device.

[Explanation of symbols]

1, 2,... First and second moving image decoding units, 3,... Motion compensation information determining unit, 4,.
Image feature analysis unit, 11 decoding unit, 14 motion compensation unit, 2
1 ... intra-frame / inter-frame coding determination unit, 26 ... motion compensation unit, 27 ... small area motion search unit, 31, 32 ... first,
Second moving image decoding unit, 33... Conversion area image synthesizing unit, 34
... Moving picture re-encoding unit.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C023 AA12 DA01 DA04 EA03 5C059 KK00 KK15 LA00 MA05 MA23 NN00 NN03 PP05 PP06 PP07 5J064 BA01 BB03 BB13 BC08 BC23 BD02 BD03

Claims (13)

[Claims] An input compression coded video data A and an input compression coded data B using inter-frame motion compensation prediction
A dissolve image encoding device that creates compressed encoded moving image data E using a dissolve effect, comprising: a moving image decoding unit that decodes the input compressed encoded moving image data A and the input compressed encoded data B; A motion compensation information determining unit that determines motion compensation information of a dissolve section from the motion compensation information output from the video decoding unit; and a decoding unit that decodes the decoded image information output from the first and second video decoding units. And a moving image re-encoding unit that re-encodes the synthesized image output from the image synthesizing unit. The motion compensation information determination unit includes the input encoded moving image data A and Using the motion compensation information of the input coded moving image data B, the motion compensation information of the compression coded moving image data E is determined, and the moving image re-encoding unit performs no motion search, and Encoding apparatus dissolve image, characterized in that the re-encoded using motion compensation information determined by the compensation information determination unit. 2. An input compression coded video data A and an input compression coded data B using inter-frame motion compensation prediction.
A dissolve image encoding device that creates compressed encoded moving image data E using a dissolve effect, comprising: a moving image decoding unit that decodes the input compressed encoded moving image data A and the input compressed encoded data B; A motion compensation information determining unit that determines motion compensation information of a dissolve section from the motion compensation information output from the video decoding unit; and a decoding unit that decodes the decoded image information output from the first and second video decoding units. And a moving image re-encoding unit that re-encodes the synthesized image output from the image synthesizing unit. The motion compensation information determination unit includes the input encoded moving image data A and Utilizing the motion compensation information of the input encoded video data B, the motion compensation information of the compression encoded video data E is determined. The constant motion compensation information as a starting point for motion search performs a motion search only in a small region around the coding apparatus dissolve image, characterized in that the re-encoded using the results. 3. The motion compensation information determination unit according to claim 1, wherein the input coded video A and the input coded video B in a dissolve section are provided.
The dissolve image encoding apparatus according to claim 1 or 2, wherein the motion compensation information having the higher input ratio is determined as the motion compensation information of the dissolve section. 4. The apparatus according to claim 1, wherein the motion compensation information determining unit determines whether the input coded video A or the input coded video B
3. The dissolve image encoding apparatus according to claim 1, wherein the motion compensation information having the larger feature amount is determined as the motion compensation information of the dissolve section. 5. The motion compensation information deciding unit determines a value reflecting a ratio between an input ratio of the input coded moving images A and B synthesized by the image synthesizing unit and a characteristic amount ratio of the input images A and B. 3. The apparatus for encoding a dissolve image according to claim 1, wherein the motion compensation information of the dissolve section is determined by using. 6. The dissolve image encoding apparatus according to claim 4, wherein pattern information of the image is used as the feature amount of the image. 7. The image pattern information includes a variance value in an image of the image, a sum of absolute errors with an average value in the image, and DCT.
7. The apparatus according to claim 6, wherein at least one of the sum of absolute values of the coefficients is used. 8. The encoding mode for re-encoding is set to an intra-screen encoding mode near the center of the dissolve section where the difference between the input ratios of two input images is small. The dissolve image encoding device according to any one of the above. 9. The dissolve image according to claim 1, wherein an encoding method of each frame in the dissolve section is the same as an encoding method of an image referring to motion compensation information. Encoding device. 10. A temporal position at which an image to be referred to for motion compensation information is switched is changed from a position obtained by calculation to a neighboring position at which the total number of one or a plurality of bidirectional prediction frames adjacent in front and behind is reduced. The encoding apparatus for a dissolve image according to claim 1, wherein: 11. After the input moving image data A and the input moving image data B are decoded up to the pixel area, a dissolve image synthesizing operation is performed, and thereafter, the motion compensation processing in the pixel area is performed based on the determined motion compensation information. The encoding device for a dissolve image according to any one of claims 1 to 10, wherein the encoding device re-encodes the image. 12. After the input moving image data A and the input moving image data B have been decoded up to the conversion area, a dissolve image synthesizing operation is performed, and then the motion compensation processing in the conversion area is performed based on the determined motion compensation information. The encoding device for a dissolve image according to any one of claims 1 to 10, wherein the encoding device re-encodes the image. 13. When the prediction frame interval differs between an image referencing the motion compensation information and an image to be re-encoded, the target motion compensation information is scaled according to a ratio of the prediction frame interval. The dissolve image encoding device according to any one of claims 1 to 11.