JPH08251582A - Encoded data editing device - Google Patents

Abstract

PURPOSE: To edit encoded data excellently without causing the input buffer of a decoder to overflow or underflow when the encoded data after editing are decoded. CONSTITUTION: An encoding part 2 generates 1st encoded data by compressing and encoding an input moving picture signal. Further, the encoding part 2 when replacing part of the 1st encoded data generates 2nd encoded data by re- encoding the input moving picture signal corresponding to the replacement place. At this time, a control part 3 makes the encoding part 2 perform the encoding by regarding a VBV(video buffering verifier) buffer occupation amount as an initial value and also controls the re-encoding by the encoding part 2 by adding dummy data, etc., for compensation so that the code amount of the 2nd encoded data matches the code amount of a substitution part 6 when the code amount of the 2nd encoded data does not exceeds the code amount of the substitution part 6 and is small. Therefore, VBV control can be performed continuously, so the data can be edited excellently without causing the decoding- side input buffer to overflow or underflow.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coded data editing apparatus suitable for editing data coded by a moving picture coding apparatus for compressing and coding a moving picture signal.

[0002]

2. Description of the Related Art In recent years, digital compression of images has been studied. In particular, various standardization proposals have been proposed for high-efficiency coding using DCT (discrete cosine transform). In the DCT, one frame is divided into a plurality of blocks (m pixels x
By dividing the video signal into frequency components in units of blocks, the redundancy in the spatial axis direction is reduced. By the way, CCITT (Intern) is used as a high-efficiency coding method for moving pictures of television signals.
ational Telegraph and Telephone Consultative Commi
As is well known, MPEG (Moving Picture Exper)
ts Group) method was proposed. In this method, not only compression by DCT (intra-frame compression) is performed within one frame, but also inter-frame compression for reducing redundancy in the time axis direction by utilizing correlation between frames is adopted. The inter-frame compression is to further reduce the bit rate by taking advantage of the fact that a general moving image is very similar to the frames before and after, and obtaining the difference between the frames before and after and encoding the difference value. . In particular, motion-compensated interframe predictive coding that reduces the prediction error by predicting the motion of an image and obtaining the interframe difference is effective.

In this motion compensation interframe predictive coding, a motion vector is obtained between the image data D (n) of the current frame and the image data D (n-1) of the previous frame. The image data obtained by decoding the encoded data of the previous frame is motion-compensated by the motion vector, and the motion-compensated reference image data D ′ (n−1) of the previous frame and the image data D (n) of the current frame are obtained. The difference value (error component due to motion prediction) is encoded and output.

Recently, a moving image compression method called MPEG2 method has been proposed which is further improved to satisfy the above-mentioned MPEG method in terms of interoperability, resolution variability and expandability. , Has already been standardized to ISO / IEC standard 13818-2. The MPEG2 system is a moving image compression coding system that incorporates the above-mentioned MPEG system (commonly referred to as MPEG1). That is, it is considered to satisfy all the applications used in the three fields of computer, broadcasting, and communication. MP
For moving picture compression encoding by the EG2 system, for example, see "ISO / IEC 13818-2 Draft Intern.
It is described in documents such as "ational Standard". That is,
In this moving image compression method, a motion compensation prediction (MC: Motion Copensation) between images and a conversion of a hybrid method in which a DCT of 8 × 8 pixels is combined are performed, and a signal obtained by this is further quantized and Apply variable length coding. Regarding types of MC prediction, forward prediction using a past image as a reference image, backward prediction using a future image as a reference image, and bidirectional (interpolation) prediction and prediction using both past and future images as reference images. There are intra modes that are not used.

This MC prediction mode can be set for each 16 × 16 pixel macrobook, but the usable mode is determined by the type of coding (picture).
The types of this picture include a picture composed only of Intra macrobooks, an I picture, a P picture composed of intra and forward prediction macrobooks, and all M pictures.
There are B pictures that allow C prediction mode, that is, there are three types of pictures.

[0006] The I picture does not use prediction but performs DCT conversion on the original image itself, and performs quantization and variable length coding. Therefore, an I picture can be composited with a single piece of encoded data, whereas a P picture can be DCT-transformed and quantized with respect to the MC prediction error signal with the past already encoded I or P picture in input image order. And variable length coding. Then, in the B picture, the DCT prediction error signal with the already coded I or P picture in the past and the future is subjected to DCT transform, quantization and variable length coding. For this reason, decoding of P and B pictures must be preceded by decoding of the reference image starting from the I picture.

In the MPEG2 system, a GOP (GROUP OF PICTURE) composed of an arbitrary number of pictures of the above type.
S)). Further, as is well known, the MPEG1 described above is also provided with a GOP, and image editing can be performed in GOP units. That is, MPEG2
In the same manner as in MPEG1, image editing can be performed in GOP units. In MPEG2, the picture coded first in a GOP is defined as an I picture, and a GOP header indicating the beginning of the GOP is inserted before the I picture. This GOP
In the header, the time code and the encoded data forming the GOP can be encoded only by the data in the GOP, that is, an independent GOP that does not refer to the image data of the previous GOP.
A closed GOP flag indicating whether or not (hereinafter, referred to as a closed GOP) is provided. Further, a broken link (broken Link) flag indicating that the GOP image data of the previous GOP must be referred to but cannot be edited is provided. In this way, in the moving image compression method of MPEG2, GO
The device is designed so that editing can be performed in units of P.

FIG. 8 is a block diagram showing an example of a conventional coded data editing apparatus provided with such an MPEG2 system coding section.

A moving image signal is input to the input terminal 1.
The moving image signal is given to the encoding unit 2 of the MPEG2 compression encoding system. The encoding unit 2 performs the encoding process of the moving image signal based on the encoding parameter control signal such as the picture type and the quantization step from the control unit 3. The encoding unit 2 also provides the control unit 3 with encoding information such as a generated code amount for controlling the code amount. The encoded data encoded by the encoding unit 2 is given to the digital signal recording units 4 and 5.

The control unit 3 gives a recording control signal to the digital signal recording units 4 and 5, respectively, and controls so that one of the digital signal recording units 4 and 5 records the encoded data. . The replacing unit 6 can read and write data in the digital signal recording units 4 and 5. For example, when the encoded data is replaced by using both digital signal recording units 4 and 5, the replacing unit 6 causes the digital signal recording units 4 and 5 to perform the replacement based on the signal from the control unit 3 which controls the replacement location. A part of the recorded one encoded data is replaced with a part of the other encoded data.

FIG. 9 is a block diagram showing a concrete example of the encoding unit 2. In FIG. 9, the original moving image signal input to the input terminal 11 is given to the coding order conversion circuit 12. The coding order conversion circuit 12 performs the original image input order and the coded image order conversion as shown in FIG. 10, and also performs the scanning order and the block / macroblock order conversion to output the coding order image signal.

FIG. 10 is an explanatory diagram for explaining the processing by the coding order conversion circuit 12 described above, and shows a structural example of the GOP and the prediction direction of each picture. The GOP structure shown in the upper part of the figure shows the input order of the original image data, that is, the image display order, and the lower part of the figure shows the GOP structure converted into the coding order after the coding order conversion. The arrows in the figure indicate the prediction direction of each picture.

In this example, one GOP is, for example, one I
It is composed of a picture, one P picture and four B pictures. Then, as shown in the upper part of the figure, the moving image signal in each picture is sequentially input. in this case,
When the coding order conversion is performed, the coding order is changed as shown in the lower part of FIG. At this time, the period between the I picture and the P picture is 3 pictures, and if the first coded GOP is GOP0, GOP3 becomes a closed GOP. In this way, the encoding order conversion circuit 12 performs each conversion process on the moving image signal to obtain the encoding order image signal composed of the moving image signal of each picture in macroblock units, and obtains this encoding order image signal. It is given to the subtraction means 13.

The subtracting means 13 predicts the predicted image signal by using the coding order image signal, thereby obtaining a prediction error signal. This prediction error signal is given to the DCT circuit 14.

The DCT circuit 14 is a two-dimensional DC for each block.
The input signal is converted into frequency components by performing T processing. This makes it possible to reduce spatial correlation components. That is, the output of the DCT circuit 14 (DCT transform coefficient) is given to the quantizing circuit 15, and the quantizing circuit 15 quantizes the DCT transform coefficient based on the quantizing step set by the quantizing control signal from the coding control circuit 26. The redundancy of the signal of one block is reduced by applying the conversion. The signal obtained by the quantization circuit 15 is VLC / grammar generating means 1.
6 and the dequantization circuit 19 of the local decoding means 29.

The VLC / grammar generating circuit 16 performs variable-length coding on the supplied quantized signal, and multiplexes an additional signal such as header data generated by the coding control circuit 26 with this to a predetermined format. Generates and outputs MPEG2 encoded data. This output encoded data is output as a final output signal via the buffer memory 17 and the output terminal 18. In addition, the VLC / grammar generation circuit 16
Supplies the code amount of the generated encoded data to the encoding control circuit 26 by a signal.

On the other hand, the inverse quantization circuit 19 of the local decoding means 29 inverses the output signal from the quantization circuit 15 based on the quantization step set by the quantization control signal from the encoding control circuit 26. Quantize. That is, this process is the reverse of the quantization performed by the quantization circuit 15. After that, the DCT transform coefficient obtained by the inverse quantization circuit 19 is given to the inverse DCT circuit 20. The inverse DCT circuit 20 has a D
A two-dimensional DCT that is the reverse of the two-dimensional DCT transformation of the CT circuit 14.
The conversion is performed, and the signal obtained by the conversion is given to the adding means 21. The adding means 21 adds the output signal from the inverse DCT circuit 29 and a motion-compensated predicted image signal described later to obtain a decoded image signal, and supplies this decoded image signal to the switch 22. That is, the local decoding means 29 outputs the decoded image signal to the switch 22 as an output signal.

The switch 22 is turned on / off by a control signal (hereinafter, picture type signal) for switching control based on the type of picture type from the encoding control circuit 26. In this case, the switch 22 operates so as to be turned off only for I and P pictures. That is, the decoded image signal is given to the reference image memory 23 via the switch 22 only for I and P pictures. At this time, the reference image memory 23 writes the given decoded image signal. Further, the reference image memory 23 always holds the decoded images of two frames encoded in the past.

The reference image memory 23 includes an encoding control circuit 2
The decoded image signal, that is, the reference image signal is controlled by 6 to read and output the stored reference image signal. At this time, the reference image memory 23
Of the reference image signals stored therein, the past and future reference image signals in the input order are given to the motion detection circuit 24. The motion detection circuit 24 detects the motion vector and the optimum prediction mode for each macroblock based on the reference image signal and the coding order image signal. At this time, the motion detection circuit 24 limits the prediction mode by detecting based on the picture type signal. The prediction mode signal obtained by the motion detection circuit 24 is given to the reference image memory 23, the motion compensation circuit 25, and the coding control circuit 26. At the same time, the motion detection circuit 24
Gives a motion vector signal corresponding to the prediction mode to the motion compensation circuit 25 and the coding control circuit 26.

The reference image memory 23 supplies a reference image signal used for prediction to the motion compensation circuit 25 based on the supplied prediction mode signal. The motion compensation circuit 25 performs motion compensation processing of the reference image signal with the prediction mode signal and the motion vector signal, and also performs interpolation processing in the time direction depending on the prediction mode to obtain a final predicted image signal. After that, this predicted image signal is given to the subtracting means 13 and the adding means 21. When the prediction mode signal is the intra mode, this prediction image signal is "0".
Therefore, since the input signal of the DCT circuit 14 is a coding order image signal, the driving image signal itself is coded without using prediction. In this way, the encoding means shown in FIG. 10 performs compression encoding of the moving image signal according to the MPEG2 system.

Parameters such as the quantization step and the picture type are input from the control unit 3 shown in FIG.
Is supplied to the encoding control circuit 26 via. Further, the coding control circuit 26 outputs the coding generation information such as the prediction mode and the generated code amount obtained at the time of coding to the output terminal 28.
It is given to the control unit 3 (see FIG. 8) via the.

By the way, the control unit 3 shown in FIG. 8 controls the code amount by changing parameters such as the quantization step given to the encoding unit 2 (see FIG. 10). In this code amount control, one is MPEG
Some are based on VBV (Video Buffering Verifier) specified in 2. Usually, coded data greatly differs depending on the characteristics of different picture types or their original images. Therefore, when the coded data is stored or transmitted in a channel having a constant bit rate, it is inevitably buffered in the encoder and the decoder. Must be provided. That is, by controlling the generated code amount by the encoding unit based on the buffer capacity on the encoder side, the code amount of the encoded data output from the buffer on the encoder side becomes a constant bit rate. In this case, information to be encoded by control based on the buffer capacity (VBV control) is added to the head portion (sequence header) of each GOP. This information is VBV and indicates the buffer capacity on the decoder side assumed by the encoder at the time of encoding. In this way, by controlling the generated code amount based on the coding information such as VBV, overflow and underflow in the buffers of the encoder and the decoder are prevented, and the actually assumed decoder is correctly I try to guarantee that it works.

As described above, in the coded data editing apparatus shown in FIG. 8, the coded data obtained by the coding section 2 is given to the digital signal recording sections 4 and 5, and based on the recording control signal from the control section 3. It is adapted to be recorded in one of the digital signal recording sections 4 and 5.

For example, when different encoded data is generated by the encoding unit 2 (see FIG. 8) and different encoded data is recorded in the digital signal recording units 4 and 5, respectively, these two encoded data It shall be edited between. Then, the digital signal recording units 4 and 5
Control signals are applied to the replacement unit 6, respectively. At the same time, the replacing section 6 gives control signals for controlling recording to the digital signal recording sections 4 and 5, respectively. That is, the replacement unit 6 can control the reading and writing of data in both digital signal recording units. In this case, the replacement unit 6 uses the control signal supplied from the control unit 3 to control the replacement location and replaces a part of the GOP of one encoded data recorded in the digital signal recording units 4 and 5 with the other. Replace with a part of GOP of encoded data. Further, the replacing unit 6 changes the broken link flag in the GOP header as needed.

Next, using the coded data editing apparatus shown in FIG. 8, only a part of the once-coded original image signal is re-encoded by different encoding parameters,
An operation example when replacing the original encoded data will be described with reference to FIG. Usually, in such re-encoding, when the encoded data obtained as the result of the first encoding is decoded, the image quality may be greatly deteriorated. Therefore, in this case, the image quality is improved by reducing the quantization step of the deteriorated part.

FIG. 11 is an explanatory diagram for explaining the operation of FIG. 8, in which the encoded data obtained by first encoding the moving image signal is the encoded data A, and a part of the moving image signal is re-encoded. The encoded data that has been encoded is referred to as encoded data B, and the encoded data that is finally obtained by editing is referred to as encoded data C.

It is assumed that a moving image signal is input to the input terminal 1. Then, the input moving image signal is transmitted to the encoding unit 2
(See FIG. 8), the above-described processing is performed and encoding is performed. At this time, the encoded data becomes the encoded data A as shown in FIG. Thereafter, the encoded data A is given to the digital signal recording unit 4 and recorded. Next, if the GOP including the portion to be re-encoded in the encoded data A is, for example, GOP1 to GOP3, the encoding unit 2 outputs the moving image signal corresponding to the GOP1 to GOP3 as the first encoded signal. Encoding is performed using an encoding parameter different from the encoding and output. At this time, since the first GOP GOP1 'cannot refer to the image of the previous GOP, the closed GOP (closed
GOP). The coded data thus obtained becomes coded data B composed of GOP1 'to GOP3' as shown in the middle part of FIG. 11, and thereafter this coded data B is given to the digital signal recording unit 5 and recorded. . Then, the replacing unit 6 replaces GOP1 to GOP3 in the encoded data A with the encoded data B, and
The broken link flag in the GOP header of No. 4 is set to "1". The data after the replacement in this way, that is, the data that has been re-encoded becomes the encoded data C as shown in the lower part of FIG. That is, GOP1 of encoded data A
It is possible to obtain encoded data C which is a result of re-encoding GOP3. Then, this encoded data C
Is recorded again by the digital signal recording unit 4, and the operation of the encoded data editing device is completed.

However, in such a coded data editing apparatus, the VBV buffer occupancy immediately before the head of the coded data B is the head of the coded data A coded first when the coded data is recoded. The VBV buffer occupancy immediately before the copy is not always the same.
In this case, the code amount of the encoded data B may not match the code amount of GOP1 to GOP3 corresponding to the replacement portion of the encoded data A. Further, the encoded data B is replaced with the replacement portion of the original encoded data A to obtain the encoded data C.
Even when is generated, after the replacement part, that is, GOP
The code amount after 4 does not match the code amount corresponding to the original encoded data A. That is, in the conventional device, the amount of generated code at the time of encoding is controlled by the information based on VBV in order to output the encoded data from the VBV buffer at a constant bit rate, but the re-encoding process accompanying the replacement work is performed. When you do
This VBV control cannot be performed normally. There is no problem when replacing all GOPs after the edited part, but when replacing only the edited part, VBV
In some cases, control cannot be performed continuously. Therefore, at the time of decoding, the input buffer of the assumed decoder may overflow or underflow.

Further, as described above, when the moving picture signal corresponding to a part of the coded data is coded, the B picture following the first I picture of the first GOP to be recoded is regarded as a past picture. This GO is highly correlated and the MC prediction mode of many macroblocks in this B picture is the forward prediction or bidirectional prediction mode.
Since P is re-encoded as a closed GOP, the code amount may increase from the original code amount, or the image quality at the time of decoding may deteriorate.

Further, a B picture following the first I picture in a GOP having a broken link flag of "1" generated by editing is not correctly decoded by the decoder, but the GOP having this flag is transmitted to the decoder. There is also a problem that the operation of the decoder when input is not specified, and thus the reproduced image varies depending on the configuration of the decoder.

[0031]

As described above, in the conventional coded data editing device, when a part of the coded data of the moving image signal is recoded and edited, the edited part in the edited coded data is edited. Since the control of the code amount by the VBV buffer is not continuously performed in the subsequent portions, the generated code amount of the replaced portion subjected to the re-encoding processing and the code amount of the encoded data corresponding to the original replacement portion match. No, after all,
It becomes impossible to output at a constant bit rate, which may cause overflow or underflow in the input buffer of the decoder at the time of decoding.

Further, when the first GOP of the moving picture signal to be re-encoded is re-encoded as a closed GOP, the code amount increases from the original code amount and the image quality at the time of decoding deteriorates. There are also points.

Further, the coded data after editing will be given to the decoder including the GOP having the broken link flag set to "1" to be decoded, but the GOP having the broken link flag will be decoded. At the time of input, since the operation of the decoder is not specified, there is also a problem that the reproduced image differs depending on the configuration of the decoder.

Therefore, the present invention has been made in view of the above problems, and edits encoded data so that the input buffer of the decoder does not overflow or underflow when the encoded data after editing is decoded. It is an object of the present invention to provide a coded data editing device capable of performing the above.

Further, according to the present invention, even when a part of the moving image signal is re-encoded to edit the encoded data, the encoding can be suppressed to the optimum code amount without increasing the code amount. Another purpose is to provide a data editing device.

Further, according to the present invention, when decoding the coded data after editing, an uncertain reproduction operation depending on the configuration of the decoder is prevented, and the reproduction image with less discomfort is edited. It is another object to provide a coded data editing device capable of performing the above.

[0037]

A coded data editing apparatus according to the present invention codes an input moving image signal to generate a coded output, and at the same time sets the state of a buffer on the decoder side assumed at the time of coding. Adopting a first code amount control for controlling a generated code amount on the basis of a state of a buffer on the side of the decoder and an encoding means for generating information including the input moving image signal to the encoding means, First encoding control means for obtaining a first encoded output by encoding using a first encoding parameter, and a part of the first encoded output as a replacement portion for generating a predetermined moving image signal The replacing means is provided to the encoding means to re-encode the state of the buffer on the decoder side at the head of the replacing section using the second encoding parameter as an initial value in the first code amount control to replace the replacing section. Output instead of the first encoded output of Together, it is obtained by including a second encoding control means for the generated code amount generated by re-encoding to match the code amount of the first substitution of the encoded output, the.

[0038]

In the present invention, the coding means codes the input moving image signal to generate a coded output, and also generates information including the state of the buffer on the decoder side assumed at the time of coding. The first coding control means adopts a first code quantity control for controlling a generated code quantity based on a state of a buffer on the decoder side, and gives an input moving image signal to the coding means to generate a first coding quantity. The first encoded output is obtained by encoding using the encoding parameter of. The second coding control means uses a part of the first coded output as a replacement section to supply a predetermined moving image signal to the coding means, and the state of the buffer on the decoder side at the head of the replacement section. Is re-encoded using the second encoding parameter as an initial value in the first code amount control, and the generated code amount generated by the re-encoding is the code amount of the replacement unit of the first encoded output. To match. As a result, since VBV control can be continuously performed at the time of encoding, encoded data can be output at a constant bit rate. Therefore, it is possible to edit the encoded data so that the edited encoded data does not overflow or underflow even when supplied to the buffer of the decoder.

[0039]

An embodiment will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the encoded data editing apparatus according to the present invention. In the device shown in FIG. 1, the same components as those of the device shown in FIG. 8 are designated by the same reference numerals.

In the present embodiment, in addition to the structure of the conventional encoded data editing apparatus shown in FIG. 8, a third digital signal recording section 7 for recording information for controlling the encoding section 2 is further provided. Is the difference.

In FIG. 1, a driving image signal is input to the input terminal 1. The moving image signal is given to the encoder 2 according to the MPEG2 system. The encoding unit 2 performs the encoding process of the moving image signal based on the encoding parameter control signal such as the picture type and the quantization step from the control unit 3 as in the conventional technique. The encoding unit 2 also provides the control unit 3 with encoding information such as a generated code amount for controlling the code amount. The encoded data encoded by the encoding unit 2 is given to the digital signal recording units 4 and 5.

The control section 3 gives a recording control signal to the digital signal recording sections 4 and 5, respectively, and controls so that one of the digital signal recording sections 4 and 5 records the encoded data. . The replacing unit 6 can read and write data in the digital signal recording units 4 and 5. For example, when the encoded data is replaced by using both digital signal recording units 4 and 5, the replacing unit 6 causes the digital signal recording units 4 and 5 to perform the replacement based on the signal from the control unit 3 which controls the replacement location. A part of the recorded one encoded data is replaced with a part of the other encoded data.

The control unit 3 also controls the coding parameters for controlling the coding unit 2 when the coding is performed, the VBV buffer occupation amount for each picture, and the generation for each picture obtained from the coding unit 2. Information such as the code amount is given to the digital signal recording unit 7. The digital signal recording section 7 records the given information. The control unit 3 can also control the reading of the information recorded in the digital signal recording unit 7. That is, the information read from the digital signal recording unit 7 is given to the control unit 3.

Next, in the case of re-encoding only a part of the moving image signal once encoded using the editing apparatus shown in FIG. 1 with different encoding parameters and replacing it with the corresponding part of the original encoded data. An operation example will be described with reference to FIG.

FIG. 2 is an explanatory diagram for explaining the operation of the encoded data editing apparatus shown in FIG. 1, and shows the GOP structure of the encoded data A obtained by encoding the moving image signal in the upper part of the figure. The middle row shows the GOP structure of the coded data B obtained by re-encoding the moving image signal at the edited location, and the lower part of the figure shows the GOP structure of the coded data C after editing. Now, it is assumed that the moving image signal is input to the input terminal 1. Then, the encoded data editing device includes the encoding unit 2
The moving image signal input by using is encoded from the beginning to the end. That is, this encoded data becomes the encoded data A shown in the upper part of the figure. Thereafter, the encoded data A is given to the digital signal recording unit 4 and recorded. At this time, information such as a coding parameter for controlling the coding unit 2 at the time of coding, the VBV buffer occupation amount for each picture, and the generated code amount for each picture is simultaneously recorded in the digital signal recording unit 7. .

Next, G including a portion to be re-encoded
The moving image signals corresponding to OPs (GOP1 to GOP3) are encoded using different encoding parameters. At this time, the control unit 3 reads from the digital signal recording unit 7 the VBV buffer occupation amount immediately before the portion to be re-encoded when the encoded data A is generated. And the control unit 3
Is used as the initial value of the VBV buffer occupancy when re-encoding the read VBV buffer occupancy, and the generated code amount of the portion that is re-encoded when the encoded data A is generated is read and the total is calculated. Then, this is set as the assigned code amount of the portion to be re-encoded. That is, the re-encoding process is performed by setting the VBV buffer occupancy used for VBV control during the re-encoding process to the same VBV buffer occupancy as the VBV buffer occupancy when the encoded data A was generated. GO at the beginning of encoded data B to be performed
The generated code amount of P is the GOP corresponding to the encoded data A.
This is to match the generated code amount of. Further, the generated code amount of the encoded data A is generated, the generated code amount of the portion to be re-encoded is read and accumulated, and the accumulated code amount is set to the allocated code amount of the portion to be re-encoded. The amount and the generated code amount of the original encoded data A of the corresponding portion are matched.

Then, the control unit 3 starts the VBV control using the set VBV buffer initial value. That is, the control unit 3 controls the encoding unit 2 so as not to exceed the assigned code amount and performs re-encoding. At this time, the coding parameters, the generated code amount, and the VBV are the same as when the first coding is performed.
The buffer occupation amount is recorded in the digital signal recording unit 7. In this way, the coded data obtained by re-encoding is coded data B (GOP1 'to GOP1' to GOP1 'through G0
OP3 '), and the encoded data B is recorded by the digital signal recording unit 5.

After that, the replacing unit 6 selects G in the encoded data A.
OP1 to GOP3 are coded data B, that is, GOP1 '
To GOP3 'are replaced. In this case, the control unit 3 gives a control signal to the replacement unit 3 to control the replacement location. Further, the control unit 3 reads out the code amount of the pictures forming the GOP1 to GOP3 and the code amount of the pictures forming the GOP1 'to GOP3' recorded in the digital signal recording unit 7, and based on these code amounts, the original The difference between the code amount and the code amount when re-encoding is calculated. For example, if, as a result, the code amount of the pictures forming GOP1 ′ to GOP3 ′ is lower than the code amount of the pictures forming GOP1 to GOP3, the control unit 3 decreases the replacement unit 6 by re-encoding. The code "0" stuffing bit (byte) is added to the encoded data B by the amount. As a result, as shown in FIG. 2, the generated code amount of the replaced portion in the coded data C after the replacement matches the generated code amount of the GOP corresponding to the original coded data A. This stuffing bit has no meaning during the decoding operation. That is, during the decoding operation, the stuffing bit following the picture data is removed from the VBV buffer together with the picture data according to the VBV standard of the MPEG2 system. Therefore, the VBV buffer occupancy immediately after the last picture data of GOP3 'is removed by the VBV buffer is the VBV buffer occupancy immediately after the last picture of GOP3 at the time of original encoding is removed from the VBV buffer. It has the same value as the occupancy. Eventually, the obtained encoded data after editing becomes the encoded data C shown in the lower part of FIG. 2, that is,
This satisfies the assumed VBV buffer capacity on the decoding side. Then, the coded data C is recorded again by the digital signal recording unit 4.

Therefore, according to the present embodiment, even if the re-encoding processing is performed for the replacement editing, the VB is surely performed.
Since the V control can be performed, the generated code amount can be controlled, and the encoded data can be output from the buffer of the encoder at a constant bit rate. Therefore, when the coded data C recorded in the edited digital signal recording unit 4 is to be decoded, the coded data C fills the VBV buffer, and therefore even if the coded data C is input to the decoder. The input buffer of the decoder can be decoded well without overflow or underflow. In order to compensate the reduction amount of the re-encoded portion, in this example, GO
Although the stuffing bit is added after P3 ', the stuffing bit is replaced by a sequence header and a sequence extension or user data immediately before the GOP header in the GOP4 after the replacement part. However, the same effect can be obtained even in this case.

In this example, the stuffing bit and the like are inserted under the control of the replacing unit 6. However, for example, the stuffing bit for compensating the code amount is inserted in the encoding unit 2. It may be configured. In this case, the data amount does not use the value recorded in the digital signal recording unit 7, but is detected from each encoded data of the digital signal recording units 4 and 5, and is set based on the detection result.

By the way, when re-encoding is performed from the middle of the moving image signal as described above, since the previous image cannot be referred to at the re-encoding start portion, the first GOP is a closed GOP. At this time, if the B picture following the first I picture in the GOP has a high correlation with the past image in the GOP, the forward coding or the bidirectional prediction is usually performed in the first coding process. It is used to reduce the code amount. However, if the re-encoding is performed as described above, the above prediction cannot be used. Therefore, if it is attempted to generate coded data capable of decoding an image having the same quality as the first coding, the code amount may increase. Further, if the coded data has a code amount similar to that of the first coding, the quality of the decoded image may be deteriorated as compared with the quality of the decoded image of the original coded data.

Therefore, when performing the first encoding, the control unit 3 refers to the degree of reference to the past image of each picture, for example, in that picture, based on the prediction mode of each macroblock obtained from the encoding unit 2. The number of macroblocks in the forward prediction and bidirectional prediction modes is recorded in the digital signal recording unit 7. When performing re-encoding, a portion of the B-picture following the first I-picture in the GOP encoding order, which has a small degree of reference to the past image, is detected by the number of the macroblocks, and re-encoding is performed for the GOP. Start from the part corresponding to. For example, the control unit 3 detects, for each image, the number of macroblocks that are predictively coded using an image before the encoded data B as a reference image, and encodes an image in which the detected number of blocks is less than a predetermined value. The GOP of the data A is set as the head of the replacement portion, and the moving image signal corresponding to this GOP is re-encoded. By this means, it is possible to suppress an increase in the code amount or deterioration of the quality of the decoded image due to re-encoding from a GOP in the middle.

In the above operation example, the reference degree of the past image for each picture is recorded in the digital signal recording unit 7 at the time of the first encoding, and the re-encoding is started later by using the reference degree. Although it has been described that the portion is determined, for example, the reference degree recorded in the digital signal recording unit 7 is
The reference level of the image of the past GOP may be used for each GOP. When the reference degree of the past picture is obtained from the encoded data by the decoder or the like, the reference degree at this time may be read and used. Furthermore, the encoding method MPEG2 used in the above apparatus is used.
In this case, since user data that can be freely used by the user can be inserted into each layer of sequence, GOP, or picture, the reference degree of the past image is described in this user data at the time of the first encoding. If you keep it
The digital signal recording unit 7 does not need to be recorded, and this information can be easily obtained from the encoded data.

Next, in the conventional technique, the GO immediately after the editing portion
Since it is necessary to set the broken link flag in the GOP header in P to "1", decoding of the B picture following the first one picture of the GOP may differ depending on the configuration of the decoder. At this time, the decoded reproduction image becomes different. Therefore, an operation for performing re-encoding to prevent such a problem will be described with reference to FIG.

FIG. 3 is an explanatory diagram for explaining another example of the operation of the apparatus shown in FIG. 1, which is a GOP in encoded data.
2 shows the relationship between the picture configuration and the prediction direction. still,
In the drawing, a portion surrounded by a thick line indicates a replacement portion to be replaced at the time of editing, and the number attached to each picture indicates the number in the input order.

Now, similarly to the conventional apparatus shown in FIG. 8, the moving image signal is encoded from the beginning to the end by using the encoding unit 2, whereby the encoded data A shown in the upper part of FIG. It is assumed that the encoded data A obtained is recorded in the digital signal recording unit 4. Here, for example, in the case of re-encoding GOP1 to GOP3 in the encoded data A with an encoding parameter different from that at the time of encoding, in the GOP next to this edited portion (GOP1 to GOP3), that is, in GOP4. The original moving picture signals corresponding to the first I picture (I26) and the following B pictures (B24, B25) are also encoded. In this case, the control unit 3
Controls the encoding unit 2 so as not to use backward prediction and bidirectional prediction for re-encoding the B pictures (B24, B25). Further, the control unit 3 controls the GOP as described above.
Regarding re-encoding of the portion corresponding to 1 to GOP3,
While controlling the code amount so that it does not exceed the original code amount,
The portion corresponding to the B picture (B24, B25) is also controlled so as not to exceed the original code amount for the same reason. As a result, the encoded data B obtained in the digital signal recording unit 5 has the picture structure and prediction direction shown in FIG.

Next, the replacing unit 6 outputs GO in the encoded data A.
P1 to GOP3 are replaced with GOP1 'to GOP3' of the encoded data B, and at the same time, B24 and B25 are replaced with B24 'and B25' in the B picture in the GOP. At this time, similarly to the above, stuffing bits for matching the code amounts before and after editing are inserted into GOP1 'to GOP3'. Similarly, for B pictures, stuffing bits and the like are added to B24 'and B25'. As a result, the coded data C after editing obtained in the digital signal recording unit 4 satisfies VBV. The coded data C has a picture structure and a prediction direction as shown in the lower part of FIG. That is,
As shown in the figure, since the GOP 4'just after the edited portion can refer to the previous GOP, the broken link flag is "0", so that the decoding process can be continuously performed. That is, it is possible to obtain encoded data that can guarantee the same decoding operation regardless of the configuration of the decoder.

In this case, contrary to the re-encoding start portion, the reference degree of the future image of each picture at the time of the first encoding, for example, the number of macroblocks for which backward prediction and bidirectional prediction are performed is previously set. It is recorded in the digital signal recording unit 7. Then, when performing re-encoding, replacement is performed by using a portion of the B-picture following the first I-picture in the GOP encoding order, which has a small reference degree of the future image, as the end of re-encoding. It is possible to suppress deterioration of image quality at the joint of the last part of the parts.

Next, another operation example of the encoded data editing apparatus shown in FIG. 1 will be described.

In the present embodiment, for example, assuming that the driving image signal is supplied to the input terminal 1 on a recording medium having high reproducibility such as a digital VTR, it is first recorded in the digital signal recording section 7. By using the parameter used in the encoding of [3] also in the re-encoding, it is possible to obtain encoded data that can refer to the image across the editing joint. An operation example in this case will be described.

Now, similarly to the conventional apparatus shown in FIG. 1, the moving image signal is encoded from the beginning to the end by using the encoding unit 2 so that the encoded data A shown in the upper part of FIG. It is assumed that the encoded data A obtained is recorded in the digital signal recording unit 4. Here, for example, when the GOP including the portion to be re-encoded is GOP1 to GOP3 of the encoded data A, re-encoding is started from the moving image signal corresponding to GOP0 which is the GOP immediately before GOP1. To do. When encoding the portion corresponding to GOP0, the control unit 3 refers to the encoding parameters when the encoded data A recorded in the digital signal recording unit 7 is generated, and at least for I and P pictures. The encoding unit 2 is controlled by the same encoding parameter. Then, re-encoding is performed using encoding parameters different from the encoding of the portion corresponding to GOP1. Also this GOP1
The code amount control is performed in the same manner as described above from the time when the coding of the portion corresponding to is started. Further, the part corresponding to the last GOP (GOP3) of the part to be re-encoded is coded other than the B picture, that is, the I and P pictures are coded using the original coding parameters. As a result, the obtained encoded data becomes the encoded data B composed of GOP0 'to GOP3' as shown in the middle part of FIG. After that, the encoded data B
Is given to the digital signal recording unit 5 for recording.

Next, the replacing unit 6 outputs GO in the encoded data A.
The operation of replacing P1 to GOP3 with GOP1 'to GOP3' of the encoded data B is performed. At this time, stuffing bits and the like are added in order to match the code amounts of the parts to be replaced, as in the above-described embodiment. This allows
The obtained encoded data becomes the encoded data C having the picture structure shown in the lower part of FIG. 4, and then recorded in the digital signal recording unit 4. Therefore, by the control described above,
I2, P5 necessary for decoding B6 'and B7' in the encoded data B and B of GOP4 in the encoded data A
I20 and P23 required to decrypt B24 and B25
Becomes the same data at the time of the first encoding and at the time of re-encoding, that is, the encoded data A and the encoded data B become the same data. That is, this means that when decoding the encoded data C after editing, the image of the previous GOP can be referred to even at the joint of editing, that is, a correct reproduced image can be obtained. In this case, I20 and P23 may not be replaced, and the original encoded data A may remain unchanged. As a result, the coded data C after editing is VB
The capacity of the V buffer is satisfied. When the reproducibility of the original moving image signal is not perfect, the picture data I2 and I5 of the encoded data A and the encoded data B of FIG. 4 are actually used.
And I20 and I23 are different from each other, so that B6 ′, B7 ′ and B24, B2 of the encoded data C are obtained.
In No. 5, decoding may not be performed correctly and image quality may deteriorate. Even in this case, the deterioration of the image quality can be suppressed by performing the above-described operation so that the portion where the degree of reference to the past image across the GOP is small is used as the joint of the editing as described above.

As described above, in the operation example of the present embodiment, the coded information is recorded in the digital signal recording unit 7 at the time of the first coding, and the digital signal recording unit 7 is used again to re-code. The encoding and editing are performed. However, when the encoding information is obtained from the encoded data, it may be read and used. Furthermore, in MPEG2, which is an encoding method used in the above apparatus, the sequence G
Since user data that can be freely used by the user can be inserted into each OP or each layer of a picture, when encoding, for example, the degree of reference of forward and backward images necessary for later editing, etc. If the coding information is described in the user data and necessary data is read from the coding information at the time of re-coding, the digital signal recording unit 7 can be omitted. is there.

FIG. 5 is a block diagram showing another embodiment of the coded data editing apparatus according to the present invention. In the device shown in FIG. 5, the same components as those of the device shown in FIG.

In this embodiment, the apparatus shown in FIG. 1 is improved, and the encoded data recorded in the digital signal recording unit 4 is provided to the encoding unit 2a.
A different point is that a coding unit 2a that performs a coding process based on the coded data is provided instead of the coding unit 2 shown in FIG.

In FIG. 5, a driving image signal is input to the input terminal 1. The moving image signal is given to the encoder 2a based on the MPEG2 system. The encoding unit 2a encodes the moving image signal based on the encoding parameter control signal such as the picture type and the quantization step from the control unit 3 as in the above embodiment. The encoding unit 2a also provides the control unit 3 with encoding information such as a generated code amount for controlling the code amount. The encoded data encoded by the encoding unit 2a is given to the digital signal recording units 4 and 5.

The control section 3 gives a recording control signal to the digital signal recording sections 4 and 5, respectively, and controls so that one of the digital signal recording sections 4 and 5 records the encoded data. . The control unit 3 also controls the reading of the encoded data recorded in the digital signal recording unit 4. Then, the read coded data read from the digital signal recording unit 4 is given to the coding unit 2a.

FIG. 6 is a block diagram showing a schematic structure of the coding unit used in FIG. 5. The same components as those of the coding unit shown in FIG. 9 are designated by the same reference numerals and the description thereof will be omitted.

In FIG. 6, the encoded data from the already encoded digital signal recording unit 4 is input to the input terminal 50. The input encoded data is given to the buffer memory 51 and stored in the buffer memory 51. The coded data is read from the buffer memory 51 in synchronization with the coded image signal in picture units, and the coded data is given to the inverse VLC circuit 52.

The inverse VLC circuit 52 decodes the parameters described in the header data of the encoded data and gives them to the encoding control circuit 26, solves the variable length code of the compressed image signal, and outputs it from the quantization circuit 15. Generate a signal that has the same data format as the signal. This signal is a quantization circuit 15 shown in FIG.
Together with the output signal of the above. The signal selection circuit 53, based on the interrupt control signal from the encoding control circuit 26, the inverse VLC circuit 52 and the quantization circuit 1
One of the five output signals is selected. The output signal of the signal selection circuit 53 is given to the VLC / grammar generation circuit 16 and the inverse quantization circuit 19. In addition, the reverse VLC circuit 52
The prediction mode signal and the motion vector signal decoded by are supplied to the signal selection circuits 54 and 55, respectively. The prediction mode signal and the motion vector signal from the motion detection circuit 24 are applied to the other input ends of these signal selection circuits 54 and 55, respectively. Then, the signal selection circuits 54 and 55 select one of the output signals of the motion detection circuit 24 and the inverse VLC circuit 52 based on the interrupt control signal and output the selected signal.

Now, assume that the interrupt control signal is a control signal indicating an interrupt mode. At this time, the signal selection circuits 53, 54 and 55 respectively obtain the signals obtained from the coded data from the outside inputted from the input terminal 50,
The output signal from the inverse VLC circuit 52 is selected and output.
In this case, the coding control circuit 26 similarly controls the picture type signal, the quantization step control signal, and the grammar generation control signal based on the parameter signal decoded by the inverse VLC circuit 52. At this time, the same encoded data as the encoded data input from the outside is reproduced as the output signal of the VLC / grammar generating circuit 16. Further, the reference image data obtained in the reference image memory 23 after the first I picture after the interruption mode is a signal obtained by decoding the encoded data supplied from the input terminal 50.

Next, using the coded data editing apparatus shown in FIGS. 5 and 6, a part of the moving image signal once coded is re-coded using different coding parameters as in the above embodiment, An operation example in the case of replacing the corresponding part of the original encoded data will be described in detail with reference to FIG. 7.

First, the moving image original signal input through the input terminal 1 is encoded by the encoding unit 2a from the beginning to the end. As a result, the encoded coded data is the encoded data A having the picture structure shown in the upper part of FIG.
Becomes After that, the encoded data A is recorded by the digital signal recording unit 4. After that, when performing re-encoding, a similar moving image signal is newly supplied from the input terminal 1 to the encoding unit 2a, and the encoded data A obtained by the first encoding is obtained from the digital signal recording unit 4. It is given to the encoding unit 2a. At this time, the encoded data A is given at a timing such that the output signal of the quantization circuit 15 and the output signal of the inverse VLC circuit 52 can be synchronized in picture units inside the encoding unit 2a in FIG.

Now, the portion to be re-encoded using different encoding parameters is the encoded data A shown in FIG.
It is assumed to be a portion corresponding to GOP1 to GOP3 in the inside. In this case, re-encoding encodes the moving image signal corresponding to the range from GOP0 to GOP3 including GOP0 immediately before it. The I of the first GOP, GOP0, the P picture, and the last GOP, the I of GOP3,
The P pictures (I20, P23) are encoded using the interrupt mode, and the other parts of GOP1 to GOP3 are controlled to encode the moving image signal as usual. The coded data obtained by this is coded data B having a picture structure as shown in the middle part of FIG. 7. Thereafter, the encoded data B is recorded by the digital signal recording section 5. The prediction direction in each picture in this case is as shown in FIG. That is, I2 and P5 in GOP0 'and I2 in GOP3'
The coded data of each picture of 0 and I23 is the same as the coded data A obtained in the first coding. Also,
The B picture (B6 ', B7') following the first I picture in the coding order in GOP1 'and each B picture in GOP3' are reference pictures obtained by decoding the coded data A obtained in the first coding. Encode using. And the replacement unit 6
7 replaces GOP1 to GOP3 of the encoded data A with GOP1 'to GOP3' of the encoded data B, as shown in FIG.
The encoded data C after editing as shown in the lower row is obtained. Then, the encoded data C is recorded in the digital signal recording unit 4.

Therefore, according to such editing, the GOP of the encoded data C finally obtained as described above is obtained.
B pictures (B6 ', B7') following the first I picture in the coding order in 1'are encoded using the reference image obtained by decoding the encoded data A obtained in the first encoding. . In addition, the coded data of each picture of I20 and I23 in GOP3 ′ used as the reference picture at the time of decoding by the B picture (B24, B25) following the first I picture in the coding order in GOP4 is the first coding. It becomes the same as the obtained encoded data A. As a result, it is possible to refer to an image across GOPs without any problem even at the joint of editing, that is, it is possible to generate encoded data that can obtain a correct decoded image.

The other effects are similar to those of the above embodiment.

Incidentally, In the present embodiment, as for the encoded data to be supplied to the encoding unit 2a, only the data of the portion in the interrupt mode is written in the buffer memory 51 in advance, and the corresponding picture data in the interrupt mode is stored in the buffer memory 51. It can also be implemented by reading out more.

Further, in this embodiment, the picture type signal and the quantization step control signal in the interrupt mode are the inverse VLC obtained from the encoded data from the input terminal 50.
It is generated based on the output signal of the circuit 52, but may be generated based on a signal from the control unit 3 of the encoded data editing device, for example.

[0080]

As described above, according to the present invention, after compression encoding of the moving image signal, a part of this compression encoded data is re-encoded and replaced with the original encoded data,
Even if the replaced encoded data after editing is given to the decoder and decoded, the encoded data can be edited so that the buffer of the decoder does not overflow or underflow.

Further, it is possible to suppress an increase in the code amount or a deterioration in the image quality at the time of decoding due to the editing of the encoded data (re-encoding from the middle of the moving image signal).

Further, there is an effect that the uncertain reproduction operation depending on the configuration of the decoder at the joint of the editing of the encoded data can be prevented and the reproduced image can be edited so that the reproduced image with less discomfort can be decoded.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of an encoded data editing device according to the present invention.

FIG. 2 is an explanatory diagram for explaining the operation of the device shown in FIG.

FIG. 3 is a diagram showing a relationship between a picture configuration and prediction of encoded data for explaining another operation example of the apparatus shown in FIG.

FIG. 4 is a diagram showing a relationship between a picture configuration of encoded data and prediction for explaining another operation example of the apparatus shown in FIG. 1.

FIG. 5 is a block diagram showing another embodiment of the encoded data editing device according to the present invention.

6 is a block diagram showing a specific configuration of an encoding unit used in the device of FIG.

7 is a diagram showing a relationship between a picture configuration of coded data and prediction for explaining the operation of the apparatus of FIG.

FIG. 8 is a block diagram showing an example of a conventional encoded data editing device.

9 is a block diagram showing a specific configuration of an encoding unit shown in FIG.

FIG. 10 is an explanatory diagram for explaining encoding according to the MPEG2 system.

11 is an explanatory diagram for explaining the operation of the apparatus in FIG.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Input terminal, 2 ... Encoding part, 3 ... Control part, 4, 5, 7
... digital signal recording section, 6 ... replacement section, 11, 18 ... output terminal, 12 ... coding order conversion circuit, 13 ... subtraction means, 1
4 ... DCT circuit, 15 ... Quantization circuit, 16 ... VLC grammar generation circuit, 19 ... Inverse quantization circuit, 20 ... Inverse DCT circuit,
21 ... Adder means, 23 ... Reference image memory, 24 ... Motion detection circuit, 25 ... Motion compensation circuit, 26 ... Encoding control circuit.

Claims (16)

[Claims] 1. A coding means for coding an input moving image signal to generate a coded output, and for generating information including a state of a buffer on the decoder side assumed at the time of coding, and the decoder side. The first code amount control for controlling the generated code amount based on the state of the buffer is adopted, the input moving image signal is given to the encoding means, and is encoded using the first encoding parameter. First encoding control means for obtaining the encoded output of the first encoding output, and a part of the first encoded output is used as a replacement portion to supply a predetermined moving image signal to the encoding means, and The state of the buffer on the decoder side is re-encoded using the second encoding parameter as an initial value in the first code amount control and output instead of the first encoded output of the replacing unit, Generated code amount generated by re-encoding Serial first encoded data editing device comprising the second encoding control means to match the code amount of substitution of the encoded output, by comprising a. 2. The second encoding control means performs re-encoding so that the generated code amount generated by the re-encoding does not exceed the code amount of the replacement unit of the first encoded output, When the generated code amount generated by the re-encoding is smaller than the code amount of the replacement unit of the first encoded output,
2. The encoding according to claim 1, wherein a difference amount between the first encoded output and the code amount of the replacement unit is obtained, and data for compensating for the difference amount is added to match the generated code amounts. Data editing device. 3. The predetermined moving image signal is an image signal of a moving image corresponding to a part of the replacement portion of the first encoded output, and the second encoding control means includes the first moving image signal. 2. The encoding according to claim 1, wherein the input moving image signal, which is the same as the image signal of the moving image corresponding to a part of the replacement part of the encoded output of, is given to the encoding means to be re-encoded. Data editing device. 4. The coded data editing apparatus according to claim 1, further comprising a recording unit for recording information including the state of the buffer on the decoder side. 5. The encoding means includes predictive encoding means for performing forward predictive encoding, backward predictive encoding, and bidirectional predictive encoding using temporally preceding and subsequent images as reference images. Of the image corresponding to the first encoded output of the replacing unit, the image predictive-coded using an image before the replacing unit as a reference image is forward-predicted at the time of re-encoding. The encoded data editing apparatus according to claim 1, wherein encoding is not adopted. 6. The encoding means includes predictive encoding means for performing forward predictive encoding, backward predictive encoding, and bidirectional predictive encoding using temporally preceding and subsequent images as reference images. Encoding control means of the first encoding output of the image corresponding to the first encoding output of the replacing unit, which is before the replacing unit.
The image predictively coded with the image No. 2 as a reference image is forward predictively coded with the second image almost the same as the first image as a reference image at the time of re-encoding. The encoded data editing device described in 1. 7. In the second image, the second encoding control means supplies the input moving image signal corresponding to the first image to the encoding means to set the first encoding parameter. The coded data editing apparatus according to claim 6, wherein the coded data editing apparatus is obtained by re-encoding using the coded data. 8. The code according to claim 6, further comprising decoding means for decoding the first encoded output corresponding to the first image to obtain the second image. Data editing device. 9. The encoded data according to claim 8, wherein the decoding means is shared with a local decoding means which the encoding means has for obtaining the reference image. Editing device. 10. The encoding means comprises predictive encoding means for performing forward predictive encoding, backward predictive encoding and bidirectional predictive encoding using temporally preceding and following images as reference images, and inputting. The encoded moving image signal is encoded in a predetermined block unit, wherein the second encoding control unit is configured to perform the encoding based on a prediction direction of a block included in the first encoded output of the replacement unit. The number of blocks predictively coded using the image before or after the replacement unit as a reference image is detected for each image, and the first encoded output corresponding to the image in which the detected number of blocks is less than a predetermined value is output as described above. The coded data editing apparatus according to claim 1, wherein the boundary is a boundary of the replacement unit. 11. A recording unit for recording a prediction direction at the time of encoding each block included in the first encoded output, wherein the second encoding control unit predicts a predetermined block from the recording unit. The direction is read, and the predictive direction and the interval between the image including the block and the reference image are used to detect that the reference image is an image prior to the replacement unit and to detect the number of the predictively coded blocks. The coded data editing device according to claim 10. 12. The encoding means is MPEG1 or M
Data to be adopted for adopting the PEG2 compression encoding method and for causing the second encoding control means to match the generated code amount generated by re-encoding with the code amount of the replacement unit of the first encoded output. 3. The coded data editing apparatus according to claim 2, wherein any one of the stuffing bit, the sequence header, and the user data is used as the above. 13. The encoding means is MPEG1 or M.
The re-encoded data adopting the compression encoding method of PEG2 and re-encoded by the second encoding control means is a series of GOPs and the first I in the GOPs following the series of GOPs.
The coded data editing device according to any one of claims 5 to 11, which is composed of a picture and a B picture following the I picture. 14. The encoding means calculates subtraction means for calculating a first prediction error signal by calculating the input moving image signal and the prediction image signal, and DCT transforms the first prediction error signal to obtain a first prediction error signal. DCT means for obtaining a DCT coefficient signal sequence of 1 and a first quantized DCT by quantizing the first DCT coefficient signal sequence
Quantization means for obtaining a coefficient signal sequence, and the first quantized DC
Variable-length coding means for variable-length coding the T coefficient sequence to obtain a coded output, and the predictive coding means performs a process reverse to that of the quantizing means to obtain a second quantized DCT coefficient signal sequence. And an inverse D that obtains a second prediction error signal by performing the inverse processing of the second quantized DCT coefficient signal sequence to that of the DCT conversion means.
CT means, adding means for adding the second prediction error signal and the predicted image signal to obtain a decoded image signal, an image memory for recording the decoded image signal, and a decoded image signal in the image memory. And a motion compensating means for motion-compensating to obtain the prediction signal, wherein the decoding means decodes a variable-length code of the encoded output to obtain a second quantized DCT coefficient signal sequence. Means and selection means for inputting the first quantized DCT coefficient signal sequence and the second quantized DCT coefficient signal sequence and selecting and outputting one of these input signals. The second encoding control means performs switching control of the selection means to dequantize one of the first quantized DCT coefficient signal sequence and the second quantized DCT coefficient signal sequence. Input to the means, the second amount Enabled when the DCT coefficient signal sequence was selected, coded data editing apparatus according to claim 8, characterized in that recording to obtain a decoded image signal of the coded output to the image memory. 15. The encoding means includes a multiplexing means, and the first encoding control means controls the multiplexing means to multiplex information based on the prediction direction as user data. The encoded data editing device according to claim 10, wherein the encoded output is generated. 16. The predetermined block unit is MPEG
The coded data editing device according to claim 10, wherein the coded data editing device is a macro block of 1 or MPEG2 system.