JP2924866B2 - Moving picture coding apparatus, moving picture decoding apparatus, moving picture coding method, and moving picture decoding method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture signal encoding apparatus, a decoding apparatus, and a moving picture signal which are used in recording, transmission, display devices, etc., and which efficiently encode a moving picture signal with a smaller code amount. The present invention relates to a signal encoding method and a signal decoding method.

[0002]

2. Description of the Related Art In high-efficiency coding of a moving image, a frame in which only a prediction error is coded by predicting a frame to be coded by a coded frame by utilizing a correlation between frames of an image signal. There is inter prediction coding. In recent years, motion-compensated inter-frame prediction, in which an image is predicted by moving the image in accordance with the motion, has become more common. However, in encoding for storage media, intra-frame independent encoding is performed without performing inter-frame prediction every few frames.
Support random access and high-speed search. FIG. 5 shows a configuration example of the encoding device.

The image signal input from the image input terminal 1 is guided to a subtractor 13 where a prediction signal coming from a changeover switch 35 is subtracted. The output of the subtractor 13, which is the prediction residual, becomes encoded data that has been encoded and compressed by the intraframe encoder 4. The data is output from the data output terminal 7 and becomes a prediction residual reproduced by the intra-frame decoder 15, and a prediction signal is added by the residual adder 16 to become a reproduced image signal. The reproduced image signal is stored in the frame memory 17, delayed by one frame, and then guided to the P terminal of the changeover switch 35.

The changeover switch 35 is controlled by a synchronizing signal separated from the input image signal, and N (4 to 15)
The movable terminal S is connected to the fixed terminal I side in an independent frame once in a frame, and is connected to the P side in other prediction frames. Here, since the reproduced image signal of the previous frame is input to P, if P is selected, inter-frame prediction is performed. Since a fixed value (0) is connected to I, the output of the subtractor 13 becomes the same as the input image signal, and the intra-frame independent encoding is performed.

FIG. 6 shows the configuration of the decoding device. The encoded data input from the data input terminal 30 is decoded by the intra-frame decoder 15, and the prediction signal supplied from the changeover switch 35 is added by the residual adder 16 to become a reproduced image signal. The signal is output from the reproduced image signal output terminal 32 and stored in the frame memory 17, and after being delayed by one frame, is input to the residual adder 16 through the changeover switch 35. The changeover switch 35 is controlled by a frame synchronization signal separated from the input data in the same manner as the encoder.

In the intra-frame encoder 4 of the encoding apparatus shown in FIG. 5, DCT is first performed, its transform output is quantized, and variable-length encoding is performed using Huffman coding or the like. The intra-frame decoder 15 of the decoding device first decodes the variable-length code, replaces the fixed-length code with a quantized representative value, and performs inverse DCT to obtain a reproduced signal. Here, the processing unit is a frame, but the same applies to a field of an interlace signal. In the following description, a decoder is also referred to as a decoder.

[0007]

In the encoding method shown as a conventional example, an independent frame is periodically provided. However, since the amount of data generated in the independent frame is larger than that in the predicted frame, the number of independent frames is increased and the number of independent frames is increased. If the functionality is to be improved, the coding efficiency will be reduced accordingly. On the other hand, the prior application by the same applicant as the present invention (Japanese Patent Application No. 3-125393, entitled "High Efficiency Coding Device and Decoding Device") solves this problem by using a frame even in an independent frame. Using inter-correlation, subtraction of the quantization error component in the encoder and addition of the inter-frame image in the decoder, while maintaining the image quality even if the quantization is coarse,
The coding efficiency has been improved.

However, in the above-described apparatus, when the pass band of the time axis LPF performed by the inter-frame addition is narrowed, the improvement effect can be improved when the image change is small. Otherwise, the image may be significantly deteriorated. Therefore, the pass band of the time axis LPF cannot be so narrowed, and as a result, the effect is reduced.

The present invention has been made in view of the above points. The inter-frame addition performed by the decoding device is also performed by the encoding device, the added signal is subtracted from the input image signal, and the residual component is encoded. By compensating for transmission, the adverse effects caused by narrowing the pass band of the time axis LPF realized by inter-frame addition can be compensated, and the pass band can be narrowed without any problem in image quality, resulting in improved coding efficiency It is an object of the present invention to provide a moving picture coding apparatus, a moving picture decoding apparatus, a moving picture coding method, and a moving picture decoding method that can reduce the amount of data.

[0010]

SUMMARY OF THE INVENTION The present invention has been made in order to achieve the above object, and is an apparatus for encoding a moving image signal, which encodes an input image signal to obtain image encoded data. Encoding means, image decoding means for decoding the image encoded data to obtain a locally decoded image signal, time axis addition means for obtaining a time axis addition signal, and subtracting the time axis addition signal from the input image signal. Subtraction means for obtaining a residual signal, a residual encoding means for encoding the residual signal to obtain residual encoded data, and decoding the residual encoded data and adding it to the time axis addition signal. Inter-picture prediction means for obtaining a locally reproduced image signal and storing the same in an image memory, wherein the time axis adding means includes a locally decoded image signal output from the image decoding means and a predicted image signal output from the image memory And the two images Were mixed with different mixing ratios depending on the goodness of fit between the issue, there is provided a video encoding apparatus to obtain the time base addition signal.

A decoding device for decoding a moving image signal encoded by the moving image encoding device, wherein the decoding device decodes the input image encoded data to obtain a decoded image signal;
A residual decoding unit that decodes the input residual encoded data to obtain a decoded residual signal, a time axis addition unit that obtains a time axis addition signal, and adds the decoded residual signal and the time axis addition signal to each other. A residual adding unit for obtaining a reproduced image signal; and an inter-image predicting unit for generating a predicted image signal from the reproduced image signal and storing the predicted image signal in an image memory, wherein the time axis adding unit is obtained by the image decoding unit. Moving image decoding in which a decoded image signal and a predicted image signal output from the image memory are mixed by changing a mixing ratio in accordance with a degree of matching between the two image signals to obtain the time-axis added signal An apparatus is provided.

Further, there is provided a method for encoding a moving image signal, comprising: an image encoding step of encoding an input image signal to obtain encoded image data; and decoding the image encoded data to obtain a locally decoded image signal. An image decoding step, a time axis addition step of obtaining a time axis addition signal, a subtraction step of subtracting the time axis addition signal from the input image signal to obtain a residual signal, and encoding the residual signal to obtain a residual signal. A residual encoding step of obtaining encoded data, and an inter-picture prediction step of decoding the residual encoded data, adding the decoded signal to the time axis addition signal, obtaining a locally reproduced image signal, and storing the signal in an image memory. The time axis adding step includes: converting a locally decoded image signal obtained in the image decoding step and a locally reproduced image signal output from the image memory in accordance with the degree of matching between the two image signals. Were mixed with different coupling ratio, it is to provide a moving picture coding method to obtain the time base addition signal.

A decoding method of a moving image signal encoded by the moving image encoding method, comprising the steps of: decoding an input encoded image data to obtain a decoded image signal; A residual decoding step of decoding the difference encoded data to obtain a decoded residual signal, a time axis adding step of obtaining a time axis added signal, and adding the decoded residual signal and the time axis added signal to obtain a reproduced image. A residual image obtaining step of obtaining a signal, and an inter-image prediction step of generating a predicted image signal from the reproduced image signal and storing the predicted image signal in an image memory. A moving image decoding method for mixing the signal and the predicted image signal output from the image memory by changing the mixing ratio in accordance with the degree of adaptation between the two image signals to obtain the time axis added signal It is intended to provide.

[0014]

The inter-frame addition performed by the decoding device is also performed by the coding device, the added signal is subtracted from the input image signal, and the residual component is encoded and transmitted. Is compensated, so that there is no problem in image quality even if the pass band of the time axis LPF realized by the inter-frame addition is narrowed. Therefore, when encoding and decoding an input image signal to obtain an image signal to be added between frames,
In the encoding, even if the quantization is coarse, the image quality can be maintained by the effect of the time axis LPF, and the amount of encoded data can be further reduced. Further, when both the predicted frame and the independent frame exist, the efficiency does not decrease so much even if the number of the independent frames is increased, so that the functionality such as the random access can be improved.

[0015]

【Example】

<First Embodiment> FIG. 1 is a block diagram showing a first embodiment of the moving picture coding apparatus according to the present invention. The same parts as those in the conventional example of FIG. In the encoder shown in FIG. 5, the encoder and the local decoder have two systems (4, 5, and 1).
4, 15, 16), and the signals of the two frames are
The difference lies in that there is a time axis adding means (9, 10, 11, 12) for changing and adding the mixture ratio in accordance with the motion judgment in the motion judgment unit 9.

In FIG. 1, an image signal input from an image signal input terminal 1 is supplied to an intra-frame encoder 4 and a subtractor 13. The intra-frame encoder 4 intra-frame encodes the input image signal, outputs encoded data from which spatial redundancy has been removed to the multiplexer 6, and also outputs the encoded data to the intra-frame decoder 5. The internal processing of the intra-frame encoder 4 is the same as in the conventional example. The intra-frame decoder 5 decodes the intra-frame coded signal, and converts the reproduced image signal of the input frame into a motion determining unit 9.
To the multiplier 10. On the other hand, the frame memory (F
M) The local reproduction image signal of another frame already decoded, which is stored in 17, is provided to the motion judgment unit 9 and the multiplier 12 as a prediction image signal. As will be described later in detail, the motion judgment unit 9 judges the suitability of the two image signals for each pixel, and provides the multipliers 10 and 12 with the mixing coefficients (1-k) and k. The multipliers 10 and 12 multiply the reproduced image signals of the two frames by (1-k) times and k times for each pixel and output to the adder 11. The adder 11 adds them, and outputs a time axis addition signal, which is the addition result, to a subtraction input terminal of the subtractor 13 and the residual adder 1.
6 and output.

FIG. 7 shows the structure of the motion judgment unit 9. A subtractor 52 calculates a difference signal between a reproduced image of the current frame supplied from the current signal input terminal 50 and a reproduced image of another frame supplied from the other signal input terminal 51. The difference signal is converted to an absolute value by an absolute value non-linear converter 53, and then non-linearly converted with a characteristic insensitive to a small difference and limited to a particularly large difference.
It is led to 4. In the spatial LPF 54, the spatial change is smoothed, and the output is gain-adjusted by the non-linear converter 55, and (1-k) and k are obtained in the range where k is 0 or more and 1 or less, and the coefficient output terminal 56 , 57. Specifically, with respect to the movement between the playback image of the current frame and the playback image of the other frame, k is about 0.2 when it is determined to be completely still, and k is when it is determined to be completely moving. If the movement is not always clear, k is a value in between. Due to the spatial LPF, k changes smoothly for each pixel independently of the DCT block of intra-frame coding. Even if the degree of matching between frames changes inside the block, it can be handled.

Here, returning to FIG. The subtractor 13
The time axis addition signal output from the adder 11 is subtracted from the input image signal and output to the intra-frame encoder 14. The intra-frame encoder 14 intra-frame encodes the residual signal, outputs the encoded data to the multiplexer 6, and also outputs the encoded data to the intra-frame decoder 15. The multiplexer 6 multiplexes the encoded data of the image signal supplied from the intra-frame encoder 4 and the encoded data of the residual signal, and outputs the encoded data to the decoding device via the data output terminal 7. The intra-frame decoder 15 decodes the intra-frame coded signal, obtains a reproduction residual signal, and supplies it to the adder 16. The adder 16 adds the reproduction residual signal and the corresponding time axis addition signal,
A locally reproduced image signal is obtained and supplied to the frame memory 17. The frame memory 17 outputs a signal obtained by delaying the locally reproduced image signal in frame units as a predicted image signal. Note that the encoding apparatus of the present embodiment does not perform the process of subtracting the quantization error of another frame before intra-frame encoding as in the above-mentioned prior application, but can perform additional processing. The image quality will be further improved.

FIG. 2 is a block diagram showing an embodiment of a decoding device corresponding to the encoding device of FIG. 1 of the present invention. 1 and 5 are denoted by the same reference numerals. In FIG. 2, a coded data signal transmitted from a coding device through a data input terminal 30 is input and demultiplexed.
Is decomposed from the multiplexed encoded data into image encoded data and residual encoded data. The encoded image data is supplied to the intra-frame decoder 5, and the residual encoded data is supplied to the intra-frame decoder 15.

The operations of the intra-frame decoders 5 and 15, the motion and determinator 9, the multipliers 10 and 12, the adders 11 and 16, and the frame memory 17 are the same as those of the encoder shown in FIG. The reproduced image signal obtained from the adder 16 is output via the reproduced image signal output terminal 32 and is also supplied to the frame memory 17. The frame memory 17 supplies a signal obtained by delaying the reproduced image signal in frame units to the motion determining unit 9 and the multiplier 12 as a predicted image signal.

According to the encoding device and the decoding device configured as described above, the inter-frame addition performed by the decoding device is also performed by the encoding device, and the added signal is subtracted from the input image signal to obtain a residual component. Is transmitted by encoding, the adverse effect of the addition processing in the decoding device is compensated. Therefore, even if the pass band of the time axis LPF realized by the inter-frame addition is narrowed, Even if the mixing ratio of frames that are temporally different from the frames is increased, there is no problem in image quality. Therefore, even if the quantization of the encoding and decoding in the intra-frame encoder 4 and the intra-frame decoder 5 is coarse, the image quality can be maintained by the effect of the time axis LPF, and the encoded data amount can be further reduced. Can be.

<Second Embodiment> FIG. 3 is a block diagram showing a second embodiment of the moving picture coding apparatus according to the present invention. FIG.
The same parts are denoted by the same reference numerals. The difference from the first embodiment shown in FIG. 1 is that only the sub-sampled low frequency components are subjected to intra-frame encoding by the intra-frame encoder 4, and the intra-frame encoding is performed for high frequency components. The inter-frame predictive coding is performed by the unit 14. Of the time axis addition signal output from the adder 21, the low frequency component is obtained by adaptively adding the output of the interpolator 8 and the output of the LPF 19, and the high frequency component is output from the motion compensator 18. Is the high frequency component of the predicted image signal. The difference between the time axis addition signal and the input image signal is encoded by the intra-frame encoder 14.

The input image signal is input to the subtractor 13 and LPF2. In the LPF 2, the input image signal is band-limited to a spatial frequency corresponding to pre-processing of the sub-sample, and is input to the sub-sampler 3. The sub-sampler 3 performs サ ブ sub-sampling both vertically and horizontally. Subsampler 3
Is input to the intra-frame encoder 4. In the intra-frame encoding 4, the intra-frame encoding is performed as in FIG. 1, and the encoded data of the low frequency component, which is the output, is transmitted to the decoding device side via the multiplexer 6 and is also transmitted to the intra-frame decoder 5. Can be The intra-frame encoder 4 operates on a subsampled image, and the number of pixels to be processed is 1/4 of the input image signal.

The intra-frame decoder 5 decodes the intra-frame coded signal, and supplies a reproduced image signal of the input frame to the interpolator 8. The interpolator 8 interpolates and creates the missing pixels in the sub-sampler 3 so as to have the same number of pixels as the original image. The output of the interpolator is provided to a motion / restoration determiner 9 and a multiplier 10. On the other hand, the local decoded image signal of another frame already decoded and stored in the frame memory (FM) 17 is input to the motion compensator 18. The motion compensator 18 detects a motion vector, and moves the image for each block according to the vector. The motion-compensated signal is input to the LPF 19 and the subtracted input terminal of the subtractor 20 as a predicted image signal. The motion vector information is also provided to the multiplexer 6.

In the LPF 19, the same band limitation as that of the LPF 2 is performed, and the output thereof is supplied to a motion judgment unit 9, a multiplier 12, and a subtraction input terminal of a subtractor 20. The operations of the motion judgment unit 9, the multipliers 10, 12, and the adder 11 are the same as those in FIG. In the subtracter 20, a low component output from the LPF 19 is subtracted from the motion compensation signal, and a motion compensated inter-frame prediction signal of a high frequency component is provided to the adder 21.

The adder 21 adds the time base addition signal of the low frequency component output from the adder 11 and the motion compensated inter-frame prediction signal of the high frequency component output from the subtracter 20. 13 subtraction input terminal and a residual adder 16. The subtractor 13 subtracts the output of the adder 21 from the input image signal to obtain a residual signal, which is provided to the intra-frame encoder 14. In the intra-frame encoder 14, the residual signal is intra-coded, and its output is supplied to the multiplexer 6 and the intra-frame decoder 15. The intra-frame encoder 15 encodes the residual with respect to the input image signal, and the number of pixels to be processed is the same as that of the input image signal.

The multiplexer 5 multiplexes two types of coded data and motion vector data and transmits the multiplexed data to the decoding device via the data output terminal 7. The residual decoded by the intra-frame decoder 15 is added to the output of the adder 21 by the adder 16 to become a locally reproduced image signal.
7, the delay is performed on a frame basis.

FIG. 4 is a block diagram showing an embodiment of a decoding apparatus corresponding to the encoding apparatus shown in FIG. 3 of the present invention. Compared to the decoding device of FIG. 2, the interpolator 8, the motion compensator 33, the LPF
19, in that a subtractor 20 is added.

The coded data coming from the data input terminal 30 is decomposed by a demultiplexer 31 into low frequency component image coded data, residual coded data, and motion vector data. To the inner decoder 5, the residual coded data is transmitted to the inner decoder 1.
5, the motion vector data is provided to the motion compensator 33.

The intra-frame decoders 5 and 15, the interpolator 8, the motion judgment unit 9, and the multiplier 1, which are the same as the local decoding unit in FIG.
0, 12, adder 11, frame memory 17, LPF1
The operations of 9, the subtractor 20, and the adders 16 and 21 are the same as those in FIG. The motion compensator 33 does not perform motion vector detection, receives only motion vector data, performs only a motion compensation operation, and outputs a predicted image signal. Since the motion vector data used is the same as that of FIG. 3, the motion-compensated signal is the same. The reproduced image signal obtained by the adder 16 is stored in the frame memory 17 and output from the reproduced image signal output 32.

With the coding apparatus and the decoding apparatus configured as described above, similarly to the first embodiment, when an input image signal is coded and decoded to obtain an image signal to be added between frames, In the encoding, even if the quantization is coarse, the image quality can be maintained by the effect of the time axis LPF, and the amount of encoded data can be further reduced. Furthermore, since only low frequency components are subjected to intra-frame independent coding, the inter-frame prediction residual of high frequency components and the time axis addition residual of low frequencies can be integrally coded, which is extremely rational. <Third Embodiment> Encoding such as the MPEG standard of ISO / IEC combining frame skipping cyclic prediction (P-frame) and non-recursive prediction (B-frame) therebetween, the first frame is encoded in the P-frame part. It is conceivable to apply the second embodiment. Although the B frame needs to be predicted from the image of the P frame, in both embodiments, the (local) reproduced image signal is stored in the frame memory 17, and the B frame can be predicted using the signal. Therefore, even if the P frame portion is changed as in the embodiment of the present invention, B is encoded and decoded in the same manner as in the prior art. On the other hand, by replacing the P frame with the first and second embodiments,
All the P frames become independent frames encoded in the frame, and have excellent functions such as random access and high-speed search.

[0032]

According to the moving picture coding apparatus and the decoding apparatus of the present invention, the interframe addition performed by the decoder is also performed by the coding apparatus, the added signal is subtracted from the input image signal, and the residual component is encoded. By transmitting the data in a coded manner, the adverse effect of the addition processing in the decoding device is compensated, so that there is no problem in image quality even if the pass band of the time axis LPF realized by the addition between frames is narrowed. Therefore, when encoding and decoding the input image signal to obtain an image signal to be added between frames, the encoding can maintain the image quality by the effect of the time-axis LPF even if the quantization is coarse. It is possible to further reduce the amount of coded data. In addition, when the independent frame includes only the low frequency component, the inter-frame prediction residual of the high frequency component and the time base addition residual of the low frequency can be integrally coded, which is reasonable. When the input image signal includes both the predicted frame and the independent frame, even if the number of the independent frames is increased, the coding efficiency does not decrease so much, and as a result, the functionality such as the random access can be improved. Furthermore, according to the present moving picture information encoding method, moving picture information having good reproduction image quality can be recorded on a recording medium for a long time, and a moving picture recording in which moving picture information encoded by the present encoding method is recorded. According to the medium, functionality such as random access and high-speed search is enhanced. As described above, the moving picture coding apparatus, the moving picture decoding apparatus, the moving picture coding method, and the moving picture decoding method of the present invention have extremely excellent practical effects.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a first embodiment of a moving picture coding apparatus according to the present invention.

FIG. 2 is a block diagram showing a first embodiment of the video decoding device of the present invention.

FIG. 3 is a block diagram showing a second embodiment of the video encoding device of the present invention.

FIG. 4 is a block diagram showing a second embodiment of the video decoding device of the present invention.

FIG. 5 is a block diagram showing a conventional example of a moving image encoding device.

FIG. 6 is a block diagram illustrating a conventional example of a video decoding device.

FIG. 7 is a diagram illustrating a configuration of a motion / evaluation device according to an embodiment.

[Explanation of symbols]

1 image signal input terminal, 2, 19 LPF, 3 subsampler, 4, 14 intraframe encoder, 5, 15 ...
In-frame decoder, 6: multiplexer, 7: data output terminal, 8: interpolator, 9: motion / evaluation determiner, 10, 12, ...
Multipliers, 11, 21, adders, 13, 20, 52, subtractors, 16 residual adders, 17 frame memories, 1
8, 33 ... motion compensator, 30 ... data input terminal, 31 ...
Demultiplexer, 32 ... reproduced image signal output terminal, 35
.., Changeover switch, 50, current signal input terminal, 51, other signal input terminal, 53, absolute value nonlinear converter, 54, spatial LPF, 55, nonlinear converter, 56, 57, coefficient output terminal.

Claims (4)

(57) [Claims] 1. An apparatus for encoding a moving image signal, comprising: an image encoding means for encoding an input image signal to obtain image encoded data; and decoding the image encoded data to obtain a locally decoded image signal. Image decoding means, time axis addition means for obtaining a time axis addition signal, subtraction means for subtracting the time axis addition signal from the input image signal to obtain a residual signal, encoding the residual signal to obtain a residual Residual encoding means for obtaining encoded data, and an inter-picture predicting means for decoding the residual encoded data, adding the decoded signal to the time axis addition signal, obtaining a local reproduction image signal, and storing the signal in an image memory, The time axis adding unit mixes the locally decoded image signal output from the image decoding unit and the predicted image signal output from the image memory by changing a mixing ratio according to a degree of matching between the two image signals. And the time axis addition signal A moving picture coding apparatus characterized by obtaining a number. 2. A decoding device for a moving image signal encoded by the moving image encoding device according to claim 1, wherein said image decoding means decodes the input image encoded data to obtain a decoded image signal. A residual decoding unit that decodes the input residual encoded data to obtain a decoded residual signal; a time axis adding unit that obtains a time axis added signal; and the decoded residual signal and the time axis added signal. A residual adding unit for obtaining a reproduced image signal by adding the image data; and an inter-image predicting unit for generating a predicted image signal from the reproduced image signal and storing the predicted image signal in an image memory; Mixing the decoded image signal obtained in step (1) with the predicted image signal output from the image memory while changing the mixing ratio in accordance with the degree of adaptation between the two image signals to obtain the time-base added signal. Moving image decoding apparatus. 3. A method for encoding a moving image signal, comprising: an image encoding step of encoding an input image signal to obtain image encoded data; and decoding the image encoded data to obtain a locally decoded image signal. An image decoding step, a time axis addition step of obtaining a time axis addition signal, a subtraction step of subtracting the time axis addition signal from the input image signal to obtain a residual signal, and encoding the residual signal to obtain a residual. A residual encoding step of obtaining encoded data; and an inter-picture prediction step of decoding the residual encoded data, adding the decoded signal to the time axis addition signal, obtaining a local reproduction image signal, and storing the signal in an image memory. The time axis adding step mixes a locally decoded image signal obtained in the image decoding step and a locally reproduced image signal output from the image memory in accordance with a degree of matching between the two image signals; A moving picture coding method characterized in that the ratio is changed and mixed to obtain the time base addition signal. 4. A method for decoding a moving image signal encoded by the moving image encoding method according to claim 3, wherein: an image decoding step of decoding input image encoded data to obtain a decoded image signal. A residual decoding step of decoding the input residual coded data to obtain a decoded residual signal; a time axis adding step of obtaining a time axis added signal; and the decoding residual signal and the time axis added signal. A residual addition step of obtaining a reproduced image signal by adding, an inter-image prediction step of generating a predicted image signal from the reproduced image signal and storing the predicted image signal in an image memory, wherein the time axis adding step is the image decoding step Mixing the decoded image signal obtained in step (1) with the predicted image signal output from the image memory while changing the mixing ratio in accordance with the degree of adaptation between the two image signals to obtain the time-base added signal. When Video decoding how.