JPH08280024A - Video signal decoder - Google Patents

Abstract

PURPOSE: To enable reverse reproduction with simple configuration using the memories with small capacity by arranging later recorded bidirectional predic tive encoder images in earlier recorded images in case of recording, and performing decoding processing by reversing the direction of prediction. CONSTITUTION: Image data, which are reproduced so that the timewise order of group-of-the-picture can be inverted, are decomposed into a B picture and an I picture inside each group of the picture by a rearranging circuit 3, the B picture recorded later at the time of recording is arranged behind the I picture recorded first, and decoding processing is performed by a decoder 4. The I picture is stored in frame memories 5 and 6. Concerning the B picture, the video signals in the frame memories 5 and 6 are sent to motion compensation circuit 9 and 10, motion compensation is performed corresponding to the predictive direction from signal switches 7 and 8, those signals are added with a video signal added and averaged by an adder 11, and converted out by a frame/ field converting circuit 14, and the I picture is outputted after being delayed for two frames.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video signal decoding apparatus for reversing and reproducing compressed image data compressed using interframe predictive coding.

[0002]

2. Description of the Related Art Conventionally, since moving image data has an extremely large amount of information, when this information is recorded for a long time, a video signal is highly efficient coded and recorded, and when this recorded signal is read out. A means for efficiently decoding is essential. In order to meet such a demand, a high-efficiency coding method utilizing the correlation of video signals has been proposed.
One is the MPEG (Moving Picture Experts Group) system.

The MPEG system first reduces the redundancy in the time axis direction by taking the difference between the image frames of the video signal by utilizing the inter-frame correlation, and then by using the line correlation, the discrete cosine transform ( The video signal is efficiently coded by reducing the redundancy in the spatial axis direction using a process such as DCT).

In the MPEG system, when the inter-frame correlation is used, the original image cannot be restored only by transmitting the difference signal between the two frame images. Pictur
e: intra-picture or intra-coded picture), P picture (Predictional Picture: forward predictive coded picture), and B picture (Bidirectionally predictive Picture: bidirectional predictive coded picture) A method of compressing and coding by combining frame images of these three pictures is used. The I picture is image data compressed only by the frame image, and the P picture is image data compressed based on the frame image and the frame image of the closest I picture before the frame image and B picture is
This data is compressed based on a total of three frame images of the frame image and the frame images of the I picture and the P picture that are closest to each other before and after the frame image. At this time, a unit when the image data for each frame is recorded and reproduced is called a group of pictures.

FIG. 4 shows a schematic configuration of an encoding device for compression-encoding a video signal and a decoding device for decoding the recorded compression-encoded data.

The video signal input from the signal input terminal 21 of FIG. 4 is compression-encoded by the encoding device 20 and recorded on the recording medium 31. Further, the compression coded data recorded on the recording medium 31 is decoded by the decoding device 40, and the reproduced image signal is output from the signal output terminal 48.

For example, in the encoding device 20, a video signal of an original image composed of a frame image for which I picture coding is designated and a frame image for which B picture coding is designated from the signal input terminal 21 is shown in A of FIG. In the case of being input for each frame image in the order shown, the video signal of the frame image for which I-picture encoding is designated bypasses the frame memory 22 and is sent to the DCT circuit 24 to be subjected to DCT processing. After the DCT coefficient is quantized by the quantization circuit 25,
The variable length coding circuit 26 performs variable length coding. This variable length coded image data is recorded on the recording medium 31.

At this time, the image data whose DCT coefficient has been quantized by the quantizing circuit 25 is also sent to the inverse quantizing circuit 27 to be inversely quantized, and after being subjected to inverse DCT processing by the IDCT circuit 28, It is written in the frame memory 29. Such processing is called local decoding.

Specifically, when the video signals of the 2I picture and the 4I picture, which are the frame images for which the two I picture encodings of FIG. 5A are designated, are locally decoded, the B picture encoding is designated. The video signal of the 3B picture which is the frame image
It is delayed by the number of frames and input to the subtractor 23.

Further, in the motion compensation circuit 30, two 2I
Each of the locally decoded 2I picture and 4I picture stored in the frame memory 29 so that the forward and backward motion-compensated predictions shown by the arrows in the upper part of FIG. The motion compensation is performed and the average thereof is taken, and then the result is output to the subtractor 23 as a motion compensation output.

In the subtractor 23, the difference between the output from the motion compensation circuit 30 and the 3B picture is taken, and the motion compensation predictive coding is performed. Thereafter, the B picture is subjected to DCT processing in the DCT circuit 24, the DCT coefficient is quantized in the quantizing circuit 25, and then the variable length coding circuit 26 performs variable length coding in the same manner as the I picture. , Is recorded on the recording medium 31. In this way, the image data string recorded on the recording medium 31, that is, the recording / reproducing data, has the order shown in B of FIG.

Here, the unit of image data surrounded by a bold line is a group of pictures. Further, to simplify the explanation, it is assumed that the group of pictures is composed of I pictures and B pictures.

The image data sequence recorded on the recording medium 31 is reproduced frame by frame and sent to the decoding device 40.

First, the I-picture image data sent to the decoding device 40 is variable-length decoded by a variable-length decoding circuit 41, dequantized by an inverse quantization circuit 42, and then I
Inverse DCT processing is performed by the DCT circuit 43 and written in the frame memory 45. The video signal of the I picture written in the frame memory 45 is delayed by 2 frames and output from the signal output terminal 48. The I-picture video signal from the IDCT circuit 43 is transferred to the frame memory 4
It is also written in 6.

Specifically, for example, when the video signals of two 2I pictures and 4I pictures are written in the frame memory 46, the image data of the 3B picture is read from the recording medium 31 and the variable length coding circuit is read. 41 is variable-length coded, dequantized by an inverse quantization circuit 42, and then inverse DCT processing is performed by an IDCT circuit 43, and an adder 44
Sent to Here, in order to perform motion compensation prediction using the two I pictures stored in the frame memory 46, the motion compensation circuit 47 performs motion compensation for each I picture, takes an average, and sends the average to the adder 44. To be In the adder 44, the video signal of the B picture from the IDCT circuit 43 and the output from the motion compensation circuit 47 are added, bypassing the frame memory 45 and the signal output terminal 4
It is output from 8. The order of the output frame images is as shown in C of FIG.

Although the frame memories 45 and 46 in the above-described decoding device are shown in a divided manner in order to explain their functions, they are actually processed by the same memory.

Further, consider the case where the reverse reproduction operation is performed in the apparatus in which the video signal is encoded and decoded as described above.

When the reverse reproduction operation is performed, the image data reproduced from the recording medium is reversely reproduced for each group of pictures in the order shown in A of FIG. 6 and input to the decoding device. The image data of each group of pictures is subjected to the rearrangement processing that is reversed for each group of pictures, and the rearrangement data 1 shown in B of FIG. 6 and the D of FIG.
The rearrangement data 2 shown in is obtained. Then, by decoding the respective image data of the rearranged data 1 and 2 of the two systems, the video signal of the reproduced image 1 shown in C of FIG. 6 and the reproduced image 2 shown in E of FIG. Get the video signal. Using these reproduced images 1 and 2, while performing the rearrangement process, the image data of the reproduced image 1 and the reproduced image 2 are switched to obtain the video signal of the image reproduced in reverse shown in F of FIG. You can

[0019]

By the way, the above-described reverse reproduction operation can be performed only in a state in which the decoding process for the image data is recorded. Therefore,
For example, in the decoding process for obtaining the rearranged data 1 of B of FIG. 6, 6I of the reproduced image data shown in C of FIG.
To obtain picture data of picture and 7B picture,
Since it is necessary to input the image data of the 5B picture and the 8I picture of B of FIG. 6 and the image data of the processing time which is twice as long as the output image data, two decoding processes are performed in parallel. Must be done. This requires two systems of decoding devices.

Further, the rearranged data 1 of FIG. 6B, FIG.
A large amount of memory is required to perform the respective rearrangement processing for obtaining the rearranged data 2 of D and the reverse reproduction image of F of FIG. In particular, a large amount of memory is required for the rearrangement process for the reverse reproduction image of F in FIG.

Further, there is a problem that the delay time from the reproduction of the image data from the recording medium to the output of the reversely reproduced video signal is long. If this delay is large, the responsiveness deteriorates when the normal reproduction operation is switched to the reverse reproduction operation in the variable speed reproduction operation or the like.

In view of the above situation, the present invention provides a video signal decoding apparatus capable of performing reverse reproduction with a simple structure using a small capacity memory.

[0023]

A video signal decoding apparatus according to the present invention is a medium in which a video signal is recorded with a set of a bidirectional predictive encoded image and an image other than the bidirectional predictive encoded image as a recording unit. From the above, the video signals are reversely reproduced so that the temporal order of each of the groups is reversed, and the bidirectional predictive coded image and the bidirectional predictive coded image inside each of the groups in the reproduced video signal that is reversely reproduced. The above problem is solved by disassembling images other than the above, disposing bidirectional predictive coded images recorded later at the time of recording in the previously recorded set, and performing the decoding process with the prediction direction in the opposite direction. .

[0024]

According to the present invention, reverse reproduction is performed from a medium on which a video signal is recorded, using a set of a bidirectional predictive coded image and an image other than the bidirectional predictive coded image as a recording unit, and the reverse reproduced video is reproduced. The bidirectional predictive coded image inside each set in the signal and the image other than the bidirectional predictive coded image are decomposed, and the bidirectional predictive coded image recorded later at the time of recording is set to the previously recorded set. The reverse reproduction operation is performed by performing rearrangement for arrangement and performing decoding processing with the prediction direction set to the reverse direction.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows a schematic configuration of a video signal decoding device according to the present invention.

A signal input terminal 1 of this video signal decoding apparatus
The image data is reproduced and input from a recording medium (not shown) so that the temporal order of the group of pictures, which is the recording unit of the image data, is reversed. The input reproduced image data is sent to the rearrangement circuit 3.

The group of pictures is composed of a B picture which is a bidirectional predictive coded image and other pictures. In this embodiment, one B picture is used.
It is assumed that it is composed of a picture and an I picture which is one intra-picture coded image.

Here, reverse signal reproduction mode information indicating whether or not reverse image reproduction operation of the reproduced image data from the signal input terminal 1 is performed is inputted from the signal input terminal 2, and the reverse image reproduction mode information is arranged. It is sent to the replacement circuit 3. When the reverse reproduction mode information indicates that the reverse reproduction operation is not performed, the rearrangement circuit 3 does not rearrange the frame images, and the image data is sent to the decoder 4 as it is.

When the reverse reproduction mode information indicates that the reverse reproduction operation is performed, the rearrangement circuit 3
The sent image data is rearranged. In this rearrangement of image data, the B picture and I picture inside each group of pictures are decomposed, and the B pictures of the group of pictures recorded later at the time of recording are changed to the group of pictures recorded before at the time of recording. Deploy.

Specifically, as shown in A of FIG. 2, when reproduced image data for each group of picture units indicated by thick lines is input, for example, 10I pictures and 9B
At the time of recording, the group of pictures consisting of pictures is recorded after the group of pictures consisting of 8I pictures and 7B pictures, and at the time of this recording, a group consisting of 10I pictures and 9B pictures recorded later in time. The 9B picture of the of picture is placed after the 8I picture in the group of pictures composed of the 8I picture and the 7B picture previously recorded at the time of recording.

In this way, the rearrangement data shown in FIG. 2B is generated.

Here, the rearranged data shown in FIG. 2B is the same as the recorded image data shown in FIG. 3B, which is the original image data shown in FIG. The image data has a similar order.

However, in the recorded image data of FIG. 3B, the prediction direction recorded together with this recorded image data, that is, the forward prediction vector, so-called forward vector, and the backward prediction vector, so-called backward vector, are in the forward direction. Since it is inserted into the image data when it is encoded, there is a problem that the definition is reversed only by rearranging the data in frame units. Further, although the direction is reversed in frame units, there is also a problem that the temporal order in the frame is reversed.

The image data rearranged by the rearrangement circuit 3 is sent to the decoder 4.

In the decoder 4, the transmitted image data is variable-length decoded and inversely quantized, and then the inverse IDCT is performed.
By performing the processing, the video signal of I picture or B picture and the prediction direction are output.

The video signal of the I picture is the frame memory 5
Is written to. After that, when the video signal of the next I picture is output from the decoder 4, the video signal of the I picture stored in the frame memory 5 is sent to the frame memory 6 subordinately connected to the frame memory 5. After being transferred, the video signal of the next I picture is transferred to the frame memory 5
Is written to.

In this way, when the video signals of the two I-picture frame images are written in the frame memories 5 and 6, the B-picture video signal is output from the decoder 4 and sent to the adder 12. At the same time, the forward vector recorded together with the image data of the B picture is the terminal a of the signal switch 7 and the terminal b of the signal switch 8.
And the backward vector is input to the terminal b of the signal switch 7 and the terminal a of the signal switch 8, respectively.

Reverse signal reproduction mode information is input to the signal switching devices 7 and 8, and the signal switching devices 7 and 8 are switched according to the reverse signal reproduction mode information. That is,
When the reverse reproduction mode information indicates that the reverse reproduction operation is not performed, the signal switches 7 and 8 are switched to the terminal b side, respectively, but when the reverse reproduction mode information indicates that the reverse reproduction operation is performed, the signal is switched. The switches 7 and 8 are switched to the terminal a side, respectively.

Therefore, when the reverse reproduction operation is performed, the signal switcher 7 switches and outputs the forward vector, and the signal switcher 8 outputs and outputs the backward vector. The forward vector from the signal switch 7 is input to the motion compensation circuit 9, and the backward vector from the signal switch 8 is input to the motion compensation circuit 10.

Here, for example, the video signal of the 10I picture of FIG. 2B is written in the frame memory 6, the video signal of the 8I picture is written in the frame memory 5, and then the video signal of the 9B picture is added by the adder 12. When the video signal of 10I picture from the frame memory 6 is sent to the motion compensating circuit 9 for forward motion compensation prediction, the video signal of 8I picture from the frame memory 5 is sent to the motion compensating circuit 10. To the motion compensation prediction in the backward direction. These motion compensated video signals are
The adder 11 adds and averages and then the adder 12
And is added to the 9B picture.

As described above, by swapping the forward vector and the backward vector and performing the motion compensation processing, the problem that the definition of the forward vector and the backward vector is reversed is solved.

The output from the adder 12 is sent to the frame / field conversion circuit 14 via the signal switch 13 switched to the terminal b side. In this frame / field conversion circuit 14, in order to solve the problem that the temporal order in the frame is reversed, the order of the fields in the frame is exchanged. This frame /
The output from the field conversion circuit 14 is output from the signal output terminal 15 as a video signal of an image reproduced in reverse as shown in C of FIG. Even in the case of a normal reproducing operation, it is necessary to convert the field order by the frame / field converting circuit 14.

The video signal of the I picture stored in the frame memories 5 and 6 and delayed by two frames is
The signal is sent to the frame / field conversion circuit 14 via the signal switch 13 switched to the terminal a side, and after the field order is exchanged, the signal output terminal 15 is sequentially output.
Output from

In the above embodiment, the 1 × speed reverse reproduction operation was described, but a lower speed reverse reproduction operation can also be performed by the processing operation in the above-mentioned video signal encoding device.

Therefore, in the video signal decoding apparatus of the above embodiment, it is possible to perform the reproduction operation at -1 to +1 times speed.

Further, if the hardware configuration is increased, it is possible to perform both the forward and backward reproduction operations at 1 × speed or more.

In this case, the reproduced image data to be input is temporarily stopped in units of frames, and the decoder is also temporarily stopped in synchronization with this, and the frame / field conversion circuit has a function. An operation to stop in field units is required.

In the above embodiment, the reverse reproduction operation of the group of picture unit image data composed of the B picture and the I picture has been described, but the group unit of picture unit image using the P picture instead of the I picture is described. It is also conceivable to perform a reverse reproduction operation of data.

[0049]

As is apparent from the above description, the video signal decoding apparatus according to the present invention uses a set of a bidirectional predictive coded image and an image other than the bidirectional predictive coded image as a recording unit. A video signal is reversely reproduced from the medium in which the signal is recorded so that the temporal order of each set is reversed, and a bidirectional predictive coded image inside each set in the reproduced video signal that is reversely reproduced and By decomposing images other than bidirectional predictive coded images, placing bidirectional predictive coded images recorded later at the time of recording in the previously recorded set, and performing decoding processing with the prediction direction in the reverse direction. It is possible to perform the reverse reproduction operation that is reversed in field units with a simple configuration using one decoding device without increasing the memory circuit of the decoding device. Further, a memory for rearranging the image data output from the decoding device is not required. Furthermore, the delay time from the input of the reproduced image data to the decoding and output of the reproduced image data is short. In addition, it is possible to easily switch between the reproduction operation in the forward direction and the reproduction operation in the reverse direction.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a video signal decoding device according to the present invention.

FIG. 2 is a diagram showing the order of images to be reproduced in reverse.

FIG. 3 is a diagram showing recorded data when encoded in a temporally opposite direction.

FIG. 4 is a schematic configuration diagram of a conventional video signal encoding device and decoding device.

FIG. 5 is a diagram showing the order of images during normal recording / reproduction.

FIG. 6 is a diagram showing an order of images during reverse reproduction.

[Explanation of symbols]

 3 Reordering circuit 4 Decoder 5, 6 Frame memory 7, 8, 13 Signal switcher 9, 10 Motion compensation circuit 11, 12 Adder 14 Frame / field conversion circuit

Claims (3)

[Claims] 1. A video in which a temporal sequence of each set is reversed from a medium on which a video signal is recorded, with a set of a bidirectional predictive encoded image and an image other than the bidirectional predictive encoded image as a recording unit. The signal is reversely reproduced, and the bidirectional predictive-coded image and the image other than the bidirectional predictive-coded image inside each of the groups in the reproduced video signal that is reversely reproduced are decomposed, and both are recorded later at the time of recording. A video signal decoding device, characterized in that a direction prediction coded image is arranged in a previously recorded set, and a decoding process is performed with the prediction direction being the reverse direction. 2. The video signal decoding apparatus according to claim 1, wherein when performing the decoding process, the order of the fields in the frame is exchanged. 3. The video signal decoding apparatus according to claim 1, wherein the image other than the bidirectional predictive coded image is an intra-coded image.