JPH08149408A - Digital animation editing method and device therefor - Google Patents

Abstract

PURPOSE: To provide an editing method/device which can edit the reproduced images with no disturbance and with small storage capacity by performing such editing processing as the cutting and connecting operations of data, etc., in a state where the coding data on the digital compression are set. CONSTITUTION: The coding data 2 on an editing object are cut at the GOP break nearest a set cassette position based on the editing information 1 including the frame numbers, etc., and these cut data 2 are stored in a memory. A coding data reconstructing part 4 connects the cut data together, and a coding data analyzing part 8 checks whether the decoded images are reproduced for the reconstructed data. Based on this checking result, a coding data preparation processing part 6 prepares the substitute data. A stuffing data preparation processing part 7 prepares the data used for control of a buffer. The coding data 9 are produced again at the part 4 after an editing operation where the total quantity of the substitute coding data and the stuffing data is added to the quantity of data occupying the buffer.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to digital moving image editing for performing editing processing such as cutting and joining on encoded data obtained by compressing digital moving image data by utilizing correlation between frames. It is about processing.

[0002]

2. Description of the Related Art In recent years, a moving picture compression coding expert proof (MPE) method has been used as a compression method for digital moving pictures.
International standardization organization called G (Moving Picture Coding Expert Group) has been working on the standardization. Currently 1.5 assuming storage media such as CD-ROM
The Mbps compression method is standardized as the first phase of the standardization (hereinafter referred to as MPEG1). Further, as a second phase, standardization work (hereinafter referred to as MPEG2) is being advanced with the aim of achieving higher image quality and application to a wide range of fields. This standardization method obtains a high compression rate by utilizing the correlation between frames.
In G1, an image quality of about 3 times VHS is compressed by about 1/100, and in MPEG2, an image quality of about S-VHS is about 20-.
This is achieved with a compression ratio of 1/30.

In recent years, moving picture coding processing and decoding processing have begun to be installed in personal computers and the like. By this MPEG encoding processing, 100
It can be compressed to one-half and can be stored in a storage device on a computer for a long time. If this digital moving image data is subjected to editing processing such as cutting and joining as encoded data, it becomes possible to perform editing work with a simple device having a relatively large storage capacity such as a personal computer.

In this compression method, each frame is processed only by an intra-frame compressed frame (hereinafter referred to as an I frame) whose compression processing is completed within the frame without referring to other frames, and a frame preceding in time. There is a unidirectional prediction frame (hereinafter, referred to as a P frame) that refers to a frame and a bidirectional prediction frame (hereinafter, referred to as a B frame) that refers to two frames temporally preceding and following, and includes one or more I frames. The encoded data is composed of a plurality of independently reproducible frames called a group of pictures (GOP).

As a prior art in the present invention, a CD-R
A description will be given below with reference to the drawings based on MPEG1 which is assumed to read data from a storage medium such as OM at a constant speed. FIG. 2 is a conceptual diagram of the MPEG1 encoding method, FIG. 3 is a diagram showing an encoded data order and a display image order, FIGS. 4 and 5 are diagrams showing a buffer state at the time of decoding, and FIG. 6 is an encoded frame. FIG. 7 is a diagram for explaining the buffer fullness indicating the occupancy ratio of the buffer corresponding to the above, and FIGS. 7 and 8 are diagrams showing the states of the buffers when cut and connected for each GOP.

In the encoding process of MPEG1, the encoding process is performed by using GOP as a processing unit capable of independent reproduction.
In FIG. 2, 12 frames are set as one GOP,
I, P, and B respectively indicate an I frame, a P frame, and a B frame, and subsequent numbers indicate the display order.
The arrow in the figure indicates the prediction direction. MPEG1
In the above, as a limitation of the structure of GOP, one or more I
There is a frame (I3 frame in Figure 2) and the encoded data must start with an I frame. Regarding the order of GOP display images, the last frame is P frames (see FIG. 2).
P12) of the above. In FIG. 2, the GOP 12
Of the frames, the display order is 2 before the I3 frame.
It indicates that there are two B frames (B1 and B2). This B frame is a bidirectional prediction frame in which the last P12 frame of the preceding GOP (n-1) and the I3 frame of GOP (n) are prediction frames, and GOP (n) and GOP ( n-1) two GO
P is required.

Regarding the encoded data structure, as shown in FIG. 3, unlike the display order, there is I3 frame encoded data first, followed by two B frame (B1, B2) encoded data. Of. The quantitative ratio of the encoded data is such that when the B frame is 1, the P frame is 2, and the I frame is 4. Such encoded data is C
In the case of D-ROM, it is usually read by the decoder at a constant transfer rate of 1.5 Mbps and sequentially decoded,
As described above, the data decoded in one frame time (1/30 seconds in the case of NTSC) differs depending on the encoding processing method (processing method of I frame, P frame, or B frame). Therefore, on the decoding processing side, there is a buffer for decoding the encoded data, the encoded data is supplied to the buffer at a constant speed, and the encoded data necessary for decoding is taken out and decoded. In FIGS. 4 and 5,
The buffer state is shown, and in the figure, the upward slope is the transfer rate from the storage medium (the slope R in the figure), and this is constant because the read transfer rate from the recording medium is constant. It shows that there is. In addition, the state of the buffer being small indicates that the encoded data is being read out from the buffer in the order of decoding processing. In this embodiment, a virtual decoder is assumed,
The coded data of each frame is read out every frame frequency (1/30 seconds in the case of NTSC). In the figure, it can be seen that a large amount of data is lost from the buffer during the decoding process of the I frame, and is relatively small during the P frame and the B frame. In actual decoding, a buffer cannot be installed indefinitely, so that encoding is performed while adding control in the encoding process so that decoding can be performed with the capacity of the buffer. Therefore, in normal encoding processing,
There is no encoded data in the buffer when you try to decode it (called buffer underflow),
In addition, the coded data used in the decoding process is too small, and the coded data read from the storage medium at a constant speed overflows from the buffer (called buffer overflow). Absent. In addition, the buffer state immediately before the encoded data of the encoded frame is read out to the decoder is expressed by time so that the buffer state will not be broken even if the encoded data is reproduced from an intermediate frame. There is bufferfulness data. The buffer fullness data is shown in FIG. In FIG. 6, the buffer fullness data of the B1 frame is indicated by Cb (B1). This is the time required for the data read from the storage medium at a constant speed from the state in which there is no data in the buffer to the buffer state immediately before the decoding of the B1 frame (b (B1) in FIG. 6). Shows.

FIG. 7 shows a case where the editing process is performed in units of GOPs and a part of the encoded data of the encoded stream A and the encoded stream B is extracted in units of GOPs. The GOP has one intra-frame coded frame, and by using this frame as a prediction frame, it is possible to reproduce most of the frames in the GOP and reproduce a series of videos. Utilizing this property, editing such as cutting and joining processing in GOP units can be performed.

[0009]

However, as shown in FIG. 2, a B frame (B1 frame and B2 frame) in which another GOP has to be decoded as a predicted frame in the GOP is preceded by an I3 frame. However, there is a problem in that when the gap between the previous GOP and the previous GOP is cut by the editing process, the reproduced images of these B frames cannot be obtained.

Further, simply connecting the coded data of the GOP may break the buffer state at the time of the above-mentioned decoding process. This is shown in FIGS. 8 and 9. The buffer state indicated by the solid line in FIG. 8 is the encoded data that has undergone the GOP-unit cut and joint processing shown in FIGS. 4 and 5, and the GOP on the left side of the figure is the GO shown in FIG.
P, and the one on the right is the GOP shown in FIG. In this way, if there is such encoded data as a result of editing, the buffer overflows and the encoded data read from the recording medium at a constant speed is not accumulated in the buffer and is discarded. I will be lost. In the buffer state shown by the solid line in FIG. 9, the left GOP is the one shown in FIG. 5, and the right GOP is the one shown in FIG. In this figure, contrary to FIG. 7, the buffer underflows, and the encoded data necessary for decoding cannot be taken out from the buffer during decoding. in this way,
However, if GOP data is simply connected, a buffer overflow or underflow will occur, and the reproduced image cannot be decoded normally.

[0011]

In a digital moving image editing method and apparatus for solving the above-mentioned problems, digital moving image compression coded data utilizing correlation between frames or fields is completely processed within a frame. Intra-frame coding frame for coding by intra-frame coding processing, unidirectional predictive coding frame for coding temporally previous frame as a prediction frame, and coding two temporally preceding and following frames as prediction frames A coding processing frame group, which is composed of coded data of a bidirectional predictive coding frame for performing the following, and has at least one frame of coded data including at least the intra-frame coding frame,
In the case where the coding processing frame group has a broken coding frame in which a frame of another coding processing frame group is a prediction frame, a coding cut processing unit that cuts the coding data in coding processing frame group units, , An encoded data creating means for creating new encoded data, a stuffing data creating means for creating stuffing data that is encoded data that does not affect a decoded image, a connecting process for a group of cut encoded processing frames, and It is composed of coded data created by the coded data creating means, and coded data reconstructing means for reconstructing the coded data including replacement of the stuffing data created by the stuffing data creating means with the coded data. To be done.

As another configuration, in addition to the above-mentioned components, a coded data analysis means for extracting data including buffer fullness data indicating the occupied amount of the buffer at the time of frame decoding from the coded data and data transfer rate data. There is also a configuration including.

[0013]

With the above structure, the present invention replaces the coded data of the first B frame in the display order of the GOP which is a broken coded frame with the coded data of which the GOP frame including the B frame is the prediction frame. For the obtained encoded data, in order to avoid the state of overflow or underflow at the time of decoding,
In the case of overflow, increase the stuffing data in the obtained encoded data by the amount of data that overflows, and in the case of underflow, the sum of the replaced encoded data and the stuffing data is larger than the original encoded data. , Reduce the amount of underflow. In addition, by combining the data amount of the buffer occupying amount of the encoded frame with the total amount of replaced encoded data and stuffing, it is possible to obtain a reproduced image without deterioration and without breaking the buffer at the time of decoding. You can

[0014]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an editing apparatus in an embodiment of the present invention, and encoded data created in the present invention will be described with reference to FIGS. 12, 13 and 14. Further, FIG. 8, FIG. 9, FIG. 10, and FIG. 11 show buffer states according to the present invention.

In FIG. 1, 1 is edit information, 2 is encoded data to be edited, 3 is an encoded data cut processing unit for cutting the encoded data in GOP units, and 4 is the cut encoded data. Connection processing based on editing information,
The coded data reconstructing unit 5 that performs processing such as replacement or addition of a part of the coded data and reconstructs the coded data temporarily stores the coded data for reconstruction. A memory unit 6 is provided, a coded data creating unit 6 creates coded data for replacing a part of the coded data, and a stuffing data 7 is coded data that does not affect the actual decoding process. A stuffing data addition processing unit, 8 is a coded data analysis processing unit for taking out data transfer rate data of input coded data or reconstructed coded data, buffer fullness data indicating a buffer occupation rate at the time of decoding, and the like. ,
Reference numeral 9 is the obtained encoded data.

In the editing process of the present invention, the cut position, the connection position, etc. of the video data recorded in the VTR are set by operating the screen of a computer such as a personal computer. The editing information also includes time axis information such as a time code and frame number, and the actual cutting and joining processing is performed based on the time information. In the coded data cutting process, the coded data is cut based on the edit information 1 at the break of the GOP closest to the set cut position of the coded data 2 to be edited. The data is once stored in the memory unit 5 such as a hard disk.
Accumulate in. The encoded data reconstruction unit 4 performs a reconstruction process that is a process of joining the cut encoded data. Then, the reconstructed coded data is processed by the coded data analysis unit 8 so that the decoded image is reproduced, whether the buffer state at the time of decoding does not overflow or underflow, or whether the buffer state between the connected GOPs is changed. Detect the difference. Based on the detection result, the encoded data creation processing unit 6 creates the encoded data for replacement, and the stuffing data creation processing unit 7 is newly created.
Then, the data for buffer adjustment is created, the coded data reconstructing unit 4 creates the coded data again, and the coded data 9 is the edited coded data 9. Hereinafter, encoded data reproduced according to the present invention will be described.

First, a B frame in a GOP is replaced with another GOP.
The coded data in which the frame inside is the prediction frame will be described. FIG. 12 shows the encoded data structure,
It shows a state in which two GOPs, which are GOPs having a structure in which a GOP frame preceded by the head B frame is used as a prediction frame and which are not continuous before editing, are connected. In addition, FIGS. 8, 9, 10, and 11 show the states of the buffers at the time of decoding. In FIGS. 8 and 9, the thick solid line is the case where the encoded data is simply connected, and the thin line is thin. The solid line is the case of encoded data corrected according to the present invention. Further, in FIGS. 10 and 11, the thin solid line indicates the simply connected encoded data, and the thin solid line indicates the buffer state of the GOP before editing,
The thick solid line shows the correction made by the present invention. In each figure, M indicates the upper limit of the buffer, slope R indicates the transfer rate, and indicates that the encoded data is supplied to the buffer at a constant transfer rate.

As the first embodiment for creating encoded data of the present invention, the first B frame of GOP (B1, B in FIG. 12) is used.
2) The I3 frame of the same GOP is used as the prediction frame,
The coded data creation processing unit 6 creates coded data in which the motion vector for each macroblock composed of 16 × 16 pixels, which is a compression processing unit, is 0 and the difference data is 0. By creating such encoded data, the image at the time of decoding becomes a reproduced image in which three identical frames continue, but there is no deterioration of the image that occurs when the prediction frame disappears. However, as is the case with the problems of the prior art, if the encoded data is directly connected, the buffer at the time of decoding may overflow or underflow. In other words, just by creating the encoded data as described above, it is almost the same as the thick solid line in FIG. 8, and there is a possibility of overflow. In order to avoid this, at least the stuffing data of the overflow amount generated in the coded data just connected is added to the coded data of the B frame. As a result, as shown by the thin solid line in the figure, the encoded data of B frame increases,
That is, a large amount of data in the buffer is read at the time of decoding the B frame, and as a result, the buffer does not overflow. In the figure, the stuffing data of the overflow amount is divided into two frames B1 and B2 and added, but it is also possible to deviate to either B frame.

Regarding the amount of stuffing data added, the last buffer fullness data b for filling the gap in the buffer state between two connected GOPs.
From (B11) and the transfer rate R, the buffer fullness data b ′ (I3) of the I3 frame is virtually obtained. The buffer fullness data of the I3 frame in the encoded data of the GOP connected to that value is obtained. When an overflow occurs, as shown in FIG. 10, b '(I3)
Is larger than b (I3). That is, after the decoding of GOP is started, there is no time variation of input / output in the buffer.
Since the buffer occupancy is high at the time of starting decoding, the buffer may overflow. In the present invention, in order to correct this state, the coded data amount of the B frame is added to the coded amount of the original B frame by b ′ ( I3) and b (I3) are added to obtain the data amount, so that the next frame (P6 in FIG. 10) is generated.
The buffer occupancy in the frame) can match the buffer fullness data of the encoded data. When underflow occurs, b (I3) is larger than b '(I3) as shown in FIG. That is,
On the contrary to the overflow, this time, the buffer occupancy is low at the start of decoding, so that the buffer may underflow. To correct this situation,
The encoded data amount of the B frame is combined with the encoded data with the I3 frame as a prediction frame and the stuffing data, and the difference between b ′ (I3) and b (I3) is subtracted from the original encoded amount of the B frame. By setting the data amount, the buffer occupation amount in the next frame (P6 frame in FIG. 11) can match the buffer fullness data of the encoded data. In the figure, the encoded data is adjusted only in the B1 frame, but even if it is performed in the B2 frame,
There is no problem if it is done with two B frames.

Next, as a second embodiment of creating encoded data, it is shown that the frame of the preceding GOP cannot be used as a prediction frame at the time of decoding the first B frame of GOP which is one of the data standardized by MPEG1. The broken link flag may be set in the encoded data. By setting this flag, the deterioration of the reproduced image is not displayed in the decoder capable of performing the skip processing of the B frame. However, the buffer status has not been improved. Therefore, if the stuffing data for the overflow data amount is added to the B frame having the broken link flag, or if there is a gap in the buffer fullness data between GOPs, the gap is filled as in the first embodiment. Therefore, by adding the stuffing data, the code amount is adjusted and the B frame encoded data is created. Further, in the case of underflow, a motion vector 0, which is the same GOP as a prediction frame, is obtained by subtracting the underflow data amount from the original encoded data amount of the B frame. , Encoded data having a difference of 0 and stuffing data.

In addition, the adjustment of the encoded data is performed by GOP
It is also possible to perform the buffer adjustment near the connection position, for example, in the last B frame (B11 in FIG. 12) of the preceding GOP. In this case as well, it is almost the same as the above-described first embodiment, and for the encoded data of this B frame
When the encoded data obtained by the connection overflows at the time of decoding, the overflow data amount is added to the encoded data, and at the time of underflow, for example, P12 in the figure.
Is used as a prediction frame, the sum of the coded data with motion vector 0 and difference 0 and the stuffing data is deleted from the original coded data by the amount of underflow,
If there is a gap in the buffer fullness data between them, the stuffing data is adjusted to fill the gap and the encoded data of the B frame is created.

Next, as a third embodiment of the encoded data, the head B frame of the GOP in FIG. 12 is deleted from the encoded data. By doing so, even with an inexpensive decoder that cannot detect the broken link flag described above, reproduced image data can be displayed as a video without deterioration. In this way, in the case where the encoded data when the B frame is deleted overflows, by adding the stuffing data to the encoded data of the B frame in the preceding GOP, as in the first embodiment, Avoid buffer overflows. In the case of underflow, the B frame data in the preceding GOP is encoded with the same GOP as the prediction frame, the motion vector is 0, and the difference data is 0, and the stuffing data is used for buffer adjustment. Is added.

The case where a frame in another GOP is coded as a prediction frame at the time of coding a frame of each GOP has been described above.
For the case where only the frames in P are predicted frames,
explain. In this case, even if the encoded data is cut for each GOP unit, if the encoded data can be reproduced,
The image does not deteriorate. However, the buffer status has not been improved.

An example of such encoded data will be described as a fourth embodiment. The structure of encoded data is shown in FIG. In this case, the I-frame is the first in the GOP encoded data order and the display order. In this case, when the encoded data obtained by the connection overflows, the frame near the connection position, in this case, not only the B frame, but also the I frame may be added, and the stuffing data of the overflow data amount is added. . Regarding the gap of the buffer fullness data between GOPs, the virtual buffer fullness data of the preceding GOP (b ′ (I3) of FIG. 10) is the buffer fullness data of the subsequent GOP (b (of FIG. 10). I
3)), then again the neighboring frames,
That is, not only for B frame but also for I frame,
The stuffing data for filling the gap amount of the buffer fullness data is added. If the buffer fullness data of the preceding GOP at the time of underflow is smaller than the buffer fullness data of the subsequent GOP, the B frame data is the same GOP as the prediction frame and the motion vector is 0 and the difference is the same as in the first embodiment. Data is 0
When the stuffing data is underflowed for buffer adjustment, the total amount of the coded data and the stuffing data should be less than the original encoded data amount by the underflow data amount. In the case of a buffer fullness gap, by making the data amount of the gap small, it is possible to obtain encoded data that does not cause a failure in the buffer state.

In the above embodiment, the cutting unit is the GOP unit for the editing process, but in actuality, the cutting process in a finer unit is desirable. G below
An example in which OP is cut in smaller units will be described as Example 5. As shown in FIG. 2, the GOP has an I-frame at the beginning and a fixed interval thereafter (in the case of FIG. 2, there are two P-frames apart from each other).
The P frame (P6, P9, P12) exists, and the B frame exists as a frame between them. In the case of such encoded data, even if the P frame ends in the middle, only the number of frames included in the GOP is changed, there is no problem in the limitation of the configuration of the GOP, and editable encoded data can be obtained. You can As described above, the above-described processing can be similarly performed on the encoded data cut in the middle of the GOP. In this way, it is possible to perform not only the GOP unit but also the editing process with more accurate encoded data as it is.

[0026]

The digital moving picture editing method and apparatus of the present invention have the following effects. (1) Since the editing process can be performed with the encoded data as it is, the editing process can be performed with a small storage capacity. (2) In the processing for processing the encoded data, the motion vector is 0,
Since it is a constant process such as difference data 0, it is possible to create encoded data without processing the actual image data, and the editing process does not require complicated hardware and is realized by software alone. It is possible.

[Brief description of drawings]

FIG. 1 is a block diagram of a digital moving image editing apparatus of the present invention.

FIG. 2 is an explanatory diagram of a video compression method.

FIG. 3 is a conceptual diagram showing a display image order and a coded data order of a moving image compression method.

FIG. 4 is a decoding buffer state diagram before editing processing.

FIG. 5: Decoding buffer state diagram before editing processing

FIG. 6 is a buffer state and buffer fullness data diagram.

FIG. 7 is a conceptual diagram of edit processing of encoded data.

FIG. 8 is a decoding buffer state of encoded data after editing and a buffer diagram after correction processing thereof.

FIG. 9 is a decoding buffer state of encoded data after editing and a buffer diagram after correction processing thereof.

FIG. 10 is a buffer diagram after correction processing of encoded data after editing.

FIG. 11 is a buffer diagram after correction processing of encoded data after editing.

FIG. 12 is an explanatory diagram of a coded data configuration example.

FIG. 13 is an explanatory diagram of a coded data configuration example.

FIG. 14 is an explanatory diagram of a coded data configuration example.

[Explanation of symbols]

 1 Editing Information 2 Coded Data 3 Coded Data Cut Processing 4 Coded Data Reconstruction Processing 5 Memory 6 Coded Data Creation Processing 7 Stuffing Data Creation Processing 8 Coded Data Analysis Processing 9 Coded Data After Editing

Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI Technical display area H04N 7/133 Z

Claims (21)

[Claims] 1. An intraframe coded frame in which digital moving picture compression coded data utilizing correlation between frames or fields is coded by an intraframe coding process in which processing is completed within a frame, temporally preceding. Unidirectional-prediction-encoded frame for encoding as a prediction frame, and bidirectional-prediction-encoded frame for encoding two frames temporally before and after as prediction frames, and at least within the frame. A broken coded frame that forms a coded frame group having one or more coded data including a coded frame, and the coded frame group uses a frame of another coded frame group as a prediction frame. In the case of having the following, an encoding cut processing for cutting the encoded data in units of encoding processing frame groups Means, coded data creation means for creating new coded data, stuffing data creation means for creating stuffing data that is coded data that does not affect the decoded image, and connection of the cut coding processing frame group. Coded data reconstructing means for performing processing and reconstructing the encoded data created by the encoded data creating means, and reconstructing the encoded data including replacing the stuffing data created by the stuffing data creating means with the encoded data. A digital moving image editing apparatus characterized by being equipped with. 2. The encoded data creating means has a motion vector of 0 and a difference data of 0 with respect to a prediction frame.
The coded data that is
Digital moving picture editing device described. 3. The coded data creation means creates coded data in which a frame of a coding processing frame group including a broken coded frame is a prediction frame as coded data of a broken coded frame, and creates the stuffing data. The means creates stuffing data in an amount larger than the amount of data overflowing the buffer during the decoding process for the encoded data created by the encoded data creating means, and the encoded data reconstructing means creates the encoded data. 3. The digital moving image editing apparatus according to claim 1, wherein the stuffing data is added to the. 4. A coded data analysis means for extracting data including buffer fullness data indicating a buffer occupation amount at the time of frame decoding and data transfer rate data from the coded data, and the coded data analysis means encodes the data. By obtaining the difference between the buffer occupancy in the processing frame group and the buffer occupancy in the consecutive coding processing frame groups, the coded data creating means predicts the frames of the coding processing frame group including the broken coding frame. The coded data to be a frame is created, and the stuffing data creating unit determines the difference between the coded data amount created by the coded data creating unit and the stuffing data amount in the buffer occupation amount of the continuous coding processing frame group. Corresponding to the coded data amount obtained by adding the coded data amount of the broken coded frame to The stuffing data is created as described above, and the coded data reconstructing means uses the coded data created by the coding creating means and the stuffing data as the coded data of the broken coded frame. Alternatively, the digital moving image editing apparatus described in 2. 5. The coded data reconstructing means adds coded data added with a broken link flag indicating that the preceding coded frame group cannot be used as a prediction frame at the time of decoding a broken coded frame. Creating the stuffing data, the stuffing data creating means creates stuffing data in an amount larger than the amount of data that overflows the buffer when decoding the encoded data, and the encoded data reconstructing means creates the stuffing data in the encoded data. 3. The method according to claim 1, wherein
Alternatively, the digital moving image editing apparatus according to claim 2. 6. A coded data analysis means for extracting data including buffer fullness data indicating a buffer occupation amount at the time of frame decoding and data transfer rate data from the coded data, and the coded data analysis means encodes the data. By obtaining the difference between the buffer occupancy in the processing frame group and the buffer occupancy in the continuous encoding processing frame group, the encoded data creating means predicts the encoding processing frame group preceding the decoding processing of the broken encoding frame. As the frame, coded data added with a broken link flag indicating that it cannot be used is created, and the stuffing data creating means creates continuous coded frames for the coded data created by the coded data creating means. Create stuffing data of the encoded data amount corresponding to the difference in the buffer occupancy of the group, Serial encoded data reconstruction unit digital video editing apparatus according to claim 1 or claim 2, wherein it is added to the stuffing data to the encoded data. 7. The coded data reconstructing means adds coded data to which a broken link flag indicating that the preceding coded frame group cannot be used as a prediction frame is added when decoding a broken coded frame. The coded data that is created and has a reduced data amount that is equal to or more than the data amount that underflows the buffer during the decoding process is the same as the broken coded frame created by the coded data creating unit. 2. A frame of a coded frame group is composed of coded data as a prediction frame and stuffing data created by the stuffing data creating means, and is used as coded data of the broken coded frame. 2. The digital moving image editing apparatus described in 2. 8. Coded data analysis means for extracting from the coded data data including buffer fullness data indicating a buffer occupation amount at the time of frame decoding and data transfer rate data, said coded data analysis means performing coding The difference between the buffer occupancy in the processing frame group and the buffer occupancy in the continuous encoding processing frame group is obtained, and the encoded data reconstructing means, when decoding the broken encoded frame, performs the preceding encoding processing frame. The group creates encoded data with a broken link flag indicating that it cannot be used as a prediction frame, and from the amount of the encoded data, the data amount of the difference in buffer occupancy of the continuous encoding processing frame group is calculated. The broken coded frame created by the coded data creating means for the coded data of the amount of coded data And a stuffing data created by the stuffing data creating means as a prediction frame using a frame of the same coding processing frame group as the prediction frame, and is used as the coded data of the broken coded frame. Item 1
Alternatively, the digital moving image editing apparatus described in 2. 9. The coded data reconstructing means adds coded data to which a preceding link frame group has been added with a broken link flag indicating that the preceding coded frame group cannot be used as a predictive frame when decoding a broken coded frame. The encoded data created by the encoded data creating means is a bidirectional data in the vicinity of the broken flag created by the encoded data creating unit. 3. The digital moving image editing apparatus according to claim 1, wherein the predictive coded frame is composed of coded data in which the same coded frame group is used as a predictive frame and stuffing data created by the stuffing data creating means. 10. Coded data analysis means for extracting from the coded data data including buffer fullness data indicating the occupied amount of the buffer at the time of frame decoding and data transfer rate data, said coded data analysis means performing coding. When the difference between the buffer occupancy in the processing frame group and the buffer occupancy in the continuous encoding processing frame group is obtained, the encoded data reconstructing means performs the preceding encoding processing frame during the decoding processing of the broken encoding frame. The group creates encoded data with a broken link flag indicating that it cannot be used as a prediction frame, and from the amount of the encoded data, the data amount of the difference in buffer occupancy of the continuous encoding processing frame group is calculated. In the vicinity of the broken flag created by the coded data creating means, the coded data of the drawn coded data amount is generated. 3. The digital moving image editing system according to claim 1, wherein the bidirectional predictive coding frame is composed of coded data in which the same coding processing frame group is used as a prediction frame and stuffing data created by the stuffing data creating means. apparatus. 11. The coded data reconstructing unit deletes a broken coded frame from the coded data, and the stuffing data creating unit exceeds the data amount that overflows a buffer during the decoding process of the coded data. 3. The digital moving image editing apparatus according to claim 1, wherein a quantity of stuffing data is created, and the coded data reconstructing means adds the stuffing data to the coded data. 12. A coding method in which the coded data reconstructing means deletes a broken coded frame from the coded data, so that the coded data has a smaller amount of data than an amount of data that underflows a buffer during a decoding process. Data is coded data in which the bidirectional predictive coded frame near the broken flag created by the coded data creating means is the same coded frame group as a predictive frame and stuffing data created by the stuffing data creating means. The digital moving image editing apparatus according to claim 1, wherein the digital moving image editing apparatus comprises: 13. An intra-frame coded frame in which digital moving picture compression-coded data utilizing correlation between frames or fields is coded by intra-frame coding processing in which processing is completed within a frame. Unidirectional-prediction-encoded frame for encoding as a prediction frame, and bidirectional-prediction-encoded frame for encoding two frames temporally before and after as prediction frames, and at least within the frame. In the case where a coding processing frame group having coded data of one frame or more including a coding frame is formed, and the coding processing frame group does not use a frame of another coding processing frame group as a prediction frame, the coding is performed. Coded cut processing means for cutting coded data in units of coding processing frames, and new coded data Encoded data generating means for generating,
Stuffing data creating means for creating stuffing data which is coded data that does not affect the decoded image, connecting processing of cut coding processing frame groups, coded data created by the coded data creating means, and stuffing A digital moving image editing apparatus comprising: an encoded data reconstructing means for reconstructing encoded data including replacement of the stuffing data created by the data creating means with the encoded data. 14. The digital moving image as claimed in claim 13, wherein the encoded data producing means produces encoded data having a motion vector of 0 and difference data of 0 with respect to a prediction frame. Editing device. 15. The stuffing data creating means creates stuffing data in an amount equal to or larger than the amount of data that overflows a buffer during the decoding process of the encoded data, and the encoded data reconstructing means creates the stuffing data in the encoded data. 15. The digital moving image editing apparatus according to claim 13, wherein data is added. 16. Coded data analysis means for extracting from the coded data data including buffer fullness data indicating a buffer occupancy amount at the time of frame decoding and data transfer rate data, the coded data analysis means is coded. The difference between the buffer occupancy in the processing frame group and the buffer occupancy in the continuous encoding processing frame group is obtained, and the stuffing data creating means corresponds to the difference in the buffer occupying amount in the continuous encoding processing frame group. 15. The digital moving image editing apparatus according to claim 13 or 14, wherein stuffing data having a coded data amount is created, and the coded data reconstructing means adds the stuffing data to the coded data. 17. Bidirectional prediction in the vicinity of the joint processing, which is created by the encoded data creating means, for encoded data in which the encoded data has a smaller amount of data than the amount of data that underflows in the buffer during the decoding process. 15. The digital moving image editing apparatus according to claim 13, wherein the coded frame is composed of coded data in which the same coded frame group is used as a prediction frame and stuffing data created by the stuffing data creating means. 18. Coded data analysis means for extracting from the coded data data including buffer fullness data indicating the occupied amount of the buffer at the time of frame decoding and data transfer rate data, said coded data analysis means coding. The difference between the buffer occupancy in the processing frame group and the buffer occupancy in the continuous encoding processing frame group is obtained, and the data of the difference in the buffer occupancy amount in the continuous encoding processing frame group is obtained from the amount of the encoded data. The coded data of the coded data amount that is subtracted from the coded data is the coded data in which the bidirectional predictive coded frame in the vicinity of the joint processing created by the coded data creating means uses the same coded frame group as the predicted frame. The data according to claim 13 or 14, characterized in that the data is composed of stuffing data created by said stuffing data creating means. Digital video editing equipment. 19. The encoding cut processing means completes the encoded data of the encoding processing frame group with an intraframe encoding frame or a unidirectional predictive encoding frame which is an intermediate frame. The digital moving image editing apparatus according to any one of 1 to 18. 20. An intra-frame coded frame in which digital moving picture compression-coded data utilizing correlation between frames or fields is coded by intra-frame coding processing in which processing is completed within a frame, temporally preceding Unidirectional-prediction-encoded frame for encoding as a prediction frame, and bidirectional-prediction-encoded frame for encoding two frames temporally before and after as prediction frames, and at least within the frame. A broken coded frame that forms a coded frame group having one or more coded data including a coded frame, and the coded frame group uses a frame of another coded frame group as a prediction frame. , The encoded data is cut in units of encoding processing frame groups, and new encoding is performed. Data, creates stuffing data that is encoded data that does not affect the decoded image, connects the cut encoding processing frame group, creates the created encoded data, and creates the created stuffing data. A digital moving image editing method characterized by performing reconstruction of encoded data including replacement with encoded data. 21. An intra-frame coded frame in which digital moving picture compression-coded data utilizing correlation between frames or fields is coded by intra-frame coding processing in which processing is completed within a frame, temporally preceding. Unidirectional-prediction-encoded frame for encoding as a prediction frame, and bidirectional-prediction-encoded frame for encoding two frames temporally before and after as prediction frames, and at least within the frame. In the case where a coding processing frame group having coded data of one frame or more including a coding frame is formed, and the coding processing frame group does not use a frame of another coding processing frame group as a prediction frame, the coding is performed. The encoded data is cut in units of encoding processing frame groups, new encoded data is created, and there is no shadow in the decoded image. And not create stuffing data is encoded data, cut encoding frames of the linkage process and the created coded data,
A digital moving image editing method, characterized in that the created stuffing data is replaced with the encoded data to reconstruct the encoded data.