JP2002300528A - Method and device for editing video stream - Google Patents

Abstract

PROBLEM TO BE SOLVED: To edit an MPEG video stream in units of not more than GOP. SOLUTION: A first part stream is cut out so that an I picture or a P picture becomes a final display picture from the MPEG video stream 1 (S1), and a second part stream is cut out so that the I or P picture becomes a head display image from the MPEG video stream 2 (S2); it is discriminated whether the head display image of the second part stream is an I picture (S3); when it is the I picture, it is left as it is (S4) and when it is a P picture, images from the I picture immediately before image the P picture to the P picture are decoded sequentially; thus, the decoded image of the P picture is obtained (S5); and image data obtained by re-encoding the decoded image of the P picture as the I picture is substituted for the P picture in a stream which is cut out (S6). Then, the first part stream and the second part stream are connected and edited, and, a stream 3 is generated (S7).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for editing an inter-frame coded video stream.
The present invention relates to a technique for reducing the scale of signal processing for editing and avoiding deterioration in image quality of a video stream after editing when editing is performed in units of OP (Group of Picture) or less.

[0002]

2. Description of the Related Art As an example of encoding a video between frames, an MPEG2 system (ISO / IEC13818-2) is used.
There is. This method includes an intra-coded image (hereinafter referred to as an I picture), an inter-coded image using unidirectional prediction (hereinafter referred to as a P picture), and an inter-coded image using bidirectional prediction (hereinafter referred to as a B picture). ) Forms an encoded stream.

FIG. 7 shows an example of a video stream encoded according to the MPEG2 system. 7A shows the display order, and FIG. 7B shows the code order. As shown in FIG. 7, an MPEG2 video stream is encoded in units of GOP. One GOP usually has 15
About the number of image data are put together. GO in FIG. 7
Each image in P♯ (n) is encoded as follows.

I2: Intra-frame coding P5: Unidirectional inter-frame predictive coding based on I2 P8: Unidirectional inter-frame predictive coding based on P5 P11: Unidirectional inter-frame predictive coding based on P8 P14: One-way interframe predictive coding based on P8 B0 and B1: P (-1) belonging to GOP♯ (n-1)
B3 and B4: Bidirectional interframe predictive coding based on I2 and P5 B6 and B7: Bidirectional interframe predictive coding based on P5 and P8 B9 and B10: Bidirectional Interframe Predictive Coding Based on P8 and P11 B12 and B13: Bidirectional Interframe Predictive Coding Based on P11 and P14 In the MPEG2 system, in an intra-coded image, an image is a block of 16 × 16 pixels ( (Hereinafter referred to as macroblock), and each macroblock is encoded using DCT transform. In the inter-frame prediction coded image, for each macroblock of the image, an arbitrary 16 ×
A method is employed in which an area having the highest correlation is searched from the 16 pixel portion (hereinafter referred to as a vector search), and the difference value is encoded using DCT transform. With regard to inter-frame predictive coding, it is necessary to increase the scale of the vector search process in order to improve the image quality with respect to the code amount.

In the MPEG2 system, when a B picture exists, the order of the encoded image and the order of the displayed image do not match. To decode a B picture, an image displayed later in the reproduction order than the image is referred to. Therefore, as shown in FIG. Will come.

[0006] Editing a video stream obtained by the MPEG2 method cannot be realized by simply extracting necessary picture data and concatenating the data. This is because the reference image of the image encoded into the P picture and the B picture is changed by the editing, so that the image cannot be normally reproduced.

For example, as shown in FIG. 8, a partial stream 1 obtained by extracting up to the eighth GOP # (m) data in an MPEG2 video stream 1 and a GOP # (n) 9 in an MPEG2 video stream 2 It is assumed that the video stream 3 is generated by simply combining the partial streams 2 from which the data after the first one is extracted.

Then, in the reproduced video, P ₍₂₎ 11 which should be decoded based on P ₍₂₎ 8 is decoded based on P ₍₁₎ 8, so that a normal image is restored. I can no longer do it. Further, the P ₍₂₎ 11 becomes to be decoded as a reference image P ₍₂₎ 14 and, P ₍₂₎ 11 and P
₍₂₎ B ₍₂₎ 9 decoded using 14 as a reference image, B ₍₂₎ 1
The images 0, B ₍₂₎ 12, B ₍₂₎ 13, B ₍₂₎ 15, and B ₍₂₎ 16 cannot be normally restored.

Therefore, when editing a video stream encoded by the MPEG2 system, it is necessary to perform editing in units of GOPs as shown in FIG. The example of FIG. 9 is an example in which the part before GOP # (1) of the video stream 1 and the part after GOP $ (r) of the video stream 2 are combined to obtain the edited stream 3. G
B0 and B1 of OP♯ (r) are GOP♯ as reference images.
P14 of (q) is used. GOP♯ (q) by editing
Since P14 is lost, it is necessary to exclude two images B0 and B1 of GOP 必要 (r) in the edited stream 3 as shown in FIG.

[0010] Also, the MPEG2 video stream is
When editing in units smaller than P, it is necessary to perform a procedure of once decoding the video stream of the edited portion to obtain a decoded video signal and then encoding again in the MPEG2 system as shown in FIG.

FIG. 10 shows an MPEG2 video stream 1
The video signal up to the 8th image of GOP # (1) and M
Video stream 3 in which video signals of the ninth image and subsequent images of GOP # (r) in PEG2 video stream 2 are combined
This is an example of generating.

In this example, first, G in video stream 1
The image of OP # (1) is decoded. Next, from the decoded video signal obtained by decoding, only the image used in the edited video stream 3 is extracted. On the other hand, as for the video stream 2, only the image used in the edited video stream 3 is extracted from the decoded video signal obtained by decoding the image of GOP♯ (r).

Subsequently, a video signal obtained by combining the image signals decoded and extracted from the streams 1 and 2 is again encoded by the MPEG2 method,
GOP♯ (reenc) is obtained. Then, a stream before GOP # (1) in video stream 1, a GOP # (rocnc) obtained by re-encoding, and video stream 2
The stream after GOP # (r) in the middle is combined to obtain the edit stream 3.

FIG. 11 shows an example of an apparatus configuration embodying the method for editing the MPEG stream shown in FIG. FIG.
Reference numeral 11 denotes a storage medium on which the video stream 1 is recorded. From this medium 11, GO in video stream 1
The image data 107 of P♯ (1) is read and input to the decoder (decoder) 13 to generate a decoded image signal 108. The image used in the edited stream 3 among the decoded images of GOP # (1) is held in the data recording device 18 such as a hard disk.

Next, the image data 101 of GOP # (r) in the video stream 2 recorded on the storage medium 12 is read out, similarly input to the decoder 13, and the decoded image signal 102 is generated. Is done. Among the decoded GOPＧ (r) images, the images used in the edit stream 3 are as follows:
The data is stored in the data recording device 18.

Next, the decoded video signals 108 and 102 held in the data recording device 18 are combined and input to the encoder 14 as a video signal 109, and a re-encoded video signal 104 is obtained. Here, since the encoder 14 also performs inter-frame encoding, the encoder 14 has a memory 20 for storing reference images, and searches for a highly correlated portion between the reference image signal 111 and the video signal 109 stored in the memory 20. And a vector search unit 19 for performing the search. Inter-frame difference signal 11 output from vector search unit 19
0 is DCT-converted by the DCT converter 15 and the signal 10
3 and the coded video signal 104 converted into the variable length code by the VLC converter 16 is obtained.

If the coded image is the next reference image, the DCT-transformed signal 103 is
The signal is inversely DCT-transformed by the T converter 21 and becomes a local decoded image signal 112, which is stored in the reference image memory 20. Stream signal 10 before GOP # (1) of video stream 1 recorded on storage medium 11
5. The re-encoded image signal 104 and the stream 106 after GOP♯ (r) in the video stream 2 recorded on the storage medium 12 are combined on the storage medium 17, thereby obtaining the edited stream 3. Becomes

[0018]

However, according to the conventional simple editing method, when editing a moving picture encoded by the MPEG method, an editing point can be selected only in GOP units. However, there is a problem that the editing position is restricted. For example, even if there is a scene change in the middle of a GOP and it is desired to edit at this point, the remaining images in the GOP must be left in the edited stream, and the intended video editing cannot be performed. It will be.

In addition, when editing a moving picture encoded by the MPEG system, a method of decoding an encoded video signal once to obtain an original image signal, and then encoding the original video signal again, requires an original video signal. Since an image including noise generated by encoding / decoding is re-encoded, there is a problem that a video signal after re-encoding is deteriorated. In particular, when performing inter-frame encoding during re-encoding, the noise generated in the reference image propagates to the inter-frame encoded image using the reference image, so that the quality of the video signal after re-encoding further deteriorates. There is a problem.

As for the size of the editing apparatus, not only an image storage means for storing a decoded image in the editing process is required, but also a re-encoding unit is required to perform re-encoding with inter-frame encoding. In the process of performing
There is a problem that the scale of the hardware device is increased and the software processing time is increased.

The present invention has been made in order to solve the above-mentioned problem.
P can be edited in units of P or less, preventing an increase in the scale of a hardware device in re-encoding processing and an increase in software processing time, and also due to the spread of noise on a reference image caused by inter-frame encoding. Provided is a video stream editing method and apparatus capable of preventing quality degradation of a video signal after re-encoding and realizing editing in units of video frames necessary for editing a video stream accompanied by an audio signal. The purpose is to:

[0022]

Means for Solving the Problems In order to solve the above problems, a video stream editing method according to the present invention has the following characteristic configuration.

(1) A video stream editing method for obtaining a continuously playable video stream by concatenating a part or all of a first video stream and a part or all of a second video stream which have been subjected to inter-frame prediction coding. A first procedure of cutting out a first partial video stream from the first video stream up to immediately before a picture coded by intra-frame prediction coding or unidirectional inter-frame prediction; and From the video stream
A second procedure for cutting out a second partial video stream after a picture coded by intra-frame coding or unidirectional inter-frame prediction, and a second procedure for cutting out the second video stream in the second procedure. A third procedure for determining whether the picture displayed immediately before the second partial stream is an intra-frame coded picture,
If it is determined in the third procedure that the picture displayed immediately before the second partial stream cut out from the second video stream is an intra-frame coded picture, this picture is 1 intra-coded picture, and if it is determined that the picture is not an intra-coded picture, a unidirectional inter-frame predictive coded picture from the intra-coded picture immediately before the picture to the picture is determined. After obtaining a decoded image of the picture by decoding in order, the decoded picture is re-encoded by intra-frame encoding, and the picture obtained by the re-encoding is defined as a first intra-coded picture. A fourth procedure for outputting, a first partial stream obtained in the first procedure, and a second partial stream obtained in the second procedure. Between the arm, characterized by comprising a fifth step of inserting the first intra-frame coded picture obtained in the fourth step.

(2) In the video stream editing method of (1), the fifth step is that after adding at least one piece of predetermined picture data immediately after the first partial video stream, The stream is generated by sequentially connecting the intra-coded picture and the second partial video stream.

(3) In the video stream editing method of (2), the picture data to be added immediately after the first partial video stream is obtained by using the last display picture of the first partial video stream as a reference picture. It is characterized in that the pictures are inter-frame predictive coded so as to display the same image.

(4) In the video stream editing method according to (1), the fifth step includes, following the first partial video stream and the first intra-coded picture, at least one After adding a picture of a predetermined bidirectional prediction type, the second partial video stream is connected.

(5) In the video stream editing method according to (4), the data of the bidirectional prediction type picture to be inserted following the first intra-coded picture is the first intra-coded picture. It is characterized in that it is a picture that has been subjected to inter-frame prediction coding so as to display the same image as the picture based on only the backward prediction to be referred to.

The video stream editing apparatus according to the present invention has the following characteristic configuration.

(6) A video stream editing apparatus that obtains a continuously reproducible video stream by concatenating a part or all of a first video stream and a part or all of a second video stream that have been subjected to inter-frame prediction coding. , A first cutout processing means for cutting out a first partial video stream from the first video stream up to immediately before a picture coded by intra-frame predictive coding or unidirectional inter-frame predictive coding; Second clipping processing means for clipping a second partial video stream from a picture coded by intra-frame encoding or unidirectional inter-frame prediction from a second video stream, and the second clipping processing means A picture displayed immediately before the second partial stream cut out from the second video stream. There a picture type determining means for determining whether a picture coded frame,
If the picture type determination means determines that the picture displayed immediately before the second partial stream cut out from the second video stream is an intra-frame coded picture, this picture is 1 intra-coded picture, and if it is determined that the picture is not an intra-coded picture, a unidirectional inter-frame predictive coded picture from the intra-coded picture immediately before the picture to the picture is determined. After obtaining a decoded image of the picture by decoding in order, the decoded picture is re-encoded by intra-frame encoding, and the picture obtained by the re-encoding is defined as a first intra-coded picture. A first intra-frame coded picture generating means for outputting, and a first partial scan obtained by the first clipping processing means. And stream, between the second portion stream obtained by the second extraction processing unit, the obtained in the first intra-frame coded picture generating means 1
And stream concatenation processing means for inserting the intra-frame coded picture.

(7) In the video stream editing apparatus of (6), the stream connection processing means adds at least one piece of predetermined picture data immediately after the first partial video stream, and The stream is generated by sequentially connecting the intra-coded picture and the second partial video stream.

(8) In the video stream editing apparatus according to (7), the picture data added immediately after the first partial video stream includes the last display picture of the first partial video stream as a reference picture, and It is characterized in that the pictures are inter-frame predictive coded so as to display the same image.

(9) In the video stream editing apparatus according to (6), the stream concatenation processing means comprises, following the first partial video stream and the first intra-coded picture, at least one After adding a picture of a predetermined bidirectional prediction type, the second partial video stream is connected.

(10) In the video stream editing apparatus according to (9), the bidirectional prediction type picture data to be inserted after the first intra-coded picture is:
The picture is characterized in that it is a picture that has been subjected to inter-frame predictive coding so as to display the same picture as the picture based on only backward prediction referring to the first intra-coded picture.

For example, in the video stream editing method and editing apparatus according to the present invention, when editing an MPEG stream, when a partial video stream is cut out from a video stream coded by inter-frame coding, the video stream 1 Then I picture or P
The video stream is cut out such that the picture becomes the final display screen, or in the video stream 2, the video stream is cut out so that the I picture or the P picture becomes the top display image.

If the first display picture at the beginning of the portion cut out from the video stream 2 is an I picture, it is left as it is. If it is a P picture, the picture from the I picture immediately before the P picture to the P picture is deleted. After the decoded picture of the P picture is obtained by sequential decoding, the image data obtained by re-encoding the decoded picture of the P picture as an I picture is replaced with the P picture in the cut-out stream.

Thereafter, the partial stream cut out from the video stream 1 and the partial stream cut out from the video stream 2 are combined to generate an edited stream.

Further, when the partial streams of the video stream 1 and the video stream 2 generated by the above method are combined, a P or B picture having the same display content as the final display picture of the partial stream 1, The display time of the edited video stream is controlled in video frame units by adding data of at least one of a B picture having the same display content as the first display picture and an I picture by fixed code.

According to the above means, an inter-frame coded video signal can be edited in a unit of GOP or less, and a picture generated at the time of re-encoding is only an intra-frame coded picture. Therefore, it is possible to prevent an increase in the scale of a hardware device in the re-encoding process and an increase in software processing time, and also to reduce the quality of a video signal after re-encoding due to the spread of noise on a reference image caused by inter-frame encoding. Deterioration can be prevented.

Further, since an image can be added to the editing result in video frame units, editing in video frame units required when editing a video stream accompanied by an audio signal or the like can be performed as a result. It can be realized.

[0040]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to FIGS.

FIG. 1 is a flow chart showing an embodiment of a stream editing method according to the present invention, in which a part or all of a first video stream and a part or a part of a second video stream which have been subjected to inter-frame predictive coding. It is assumed that an editing process is performed to obtain a video stream that can be continuously played back by connecting all of them.

First, a first partial video stream is extracted from the first video stream up to immediately before a picture coded by intra-frame prediction coding or unidirectional inter-frame prediction (S1). Next, a second partial video stream is cut out from the second video stream after the picture encoded by intra-frame encoding or unidirectional inter-frame prediction (S2).

Here, whether the picture displayed immediately before the second partial stream cut out from the second video stream in step S2 is an intra-frame coded picture (I picture in the case of the MPEG system) It is determined whether or not it is (S3).

If it is determined in step S3 that the picture displayed immediately before the second partial stream cut out from the second video stream is an intra-frame coded picture, this picture is First
(S4).
When it is determined in step S3 that the picture is not an intra-frame coded picture, the unidirectional inter-frame predictive coded picture from the intra-frame coded picture immediately before the picture to the picture is sequentially decoded. Thus, a decoded image of the picture is obtained (S5). Then, the decoded picture is re-encoded by intra-frame encoding, and the picture obtained by this
(S6).

Finally, between the first partial stream obtained in step S1 and the second partial stream obtained in step S2, the first intra-frame encoding obtained in step S4 or S6 A picture is inserted, and a concatenation process is performed (S7). As a result, a video stream capable of continuously reproducing the first partial stream and the second partial stream is obtained.

FIG. 2 is a block diagram showing a configuration of a video stream editing apparatus according to the above-mentioned editing method. In FIG. 2, the first cutout processing unit 31 performs processing from the first video stream stored in the storage medium A1 to immediately before a picture coded by intra-frame prediction coding or unidirectional inter-frame prediction. , Cut out the first partial video stream. Similarly, the second cutout processing unit 32
Extracts a second partial video stream after a picture encoded by intra-frame encoding or unidirectional inter-frame prediction from the second video stream stored in the storage medium A2.

The picture type judging section 33 converts the picture displayed immediately before the second partial stream cut out from the second video stream by the second cut-out processing section 32 into an intra-frame coded picture (MPEG format). It is determined whether or not the picture is an I-picture). This determination result is sent to the stream connection processing unit 34.

When the picture type determination section 33 determines that the picture displayed immediately before the second partial stream is an intra-frame coded picture, the stream connection processing section 34 It is assumed to be one intra-frame coded picture. If it is determined that the picture is not an intra-frame coded picture, the unidirectional inter-frame predictive coded picture from the intra-frame coded picture immediately before the picture to the picture is sequentially decoded. After obtaining a decoded image of a picture, re-encoding the decoded picture by intra-frame encoding,
The picture obtained by this re-encoding is defined as a first intra-coded picture. Then, the first intra-coded picture is inserted between the first partial stream and the second partial stream to perform a concatenation process. As a result, a video stream capable of continuously reproducing the first partial stream and the second partial stream is obtained, and the storage medium A
3 can be stored.

The video stream editing method and apparatus having the above configuration will be described with reference to a specific embodiment, assuming that a video stream of the MPEG system is edited.

In the embodiment shown in FIG. 3, GOP # (1) in video stream 1
OP♯ (1 ′) is generated. When making this cutout,
The stream is extracted so that the last picture in the display order is an I picture or a P picture, that is, in a code order, at a position immediately before an arbitrary I picture or a P picture. FIG.
In the example of (1), data up to P8 in the display order (up to immediately before P11 in the code order) is extracted from the data of GOP # (1), and is extracted as GOP # (1 ').

Next, GOP♯ (x) in video stream 2
Extract a part from the end. When performing this extraction, a stream is extracted at a position from an I picture or a P picture in code order. In the example of FIG. 3, data after P11 is extracted from GOP♯ (r) in code order.

At this time, when the first picture in code order among the extracted data is a P picture, or when the extracted data includes a B picture to be reproduced earlier than the first picture in code order. , Or both, a reference image for decoding these pictures is required.

Therefore, in this embodiment, first, an I picture or a P picture immediately before the extracted portion is searched.
If this is an I-picture, the data of the I-picture is extracted as it is and combined immediately before the previously cut out portion to make GOP♯ (r ′). If the picture is a P picture, I and P pictures are decoded in order from the closest I picture, and a decoded picture of the P picture is obtained.

In the example of FIG. 3, from the original GOP♯ (r)
2, P5 and P8 are decoded in order to obtain a decoded image of P8. An I picture (I8 in the example of FIG. 3) obtained by re-encoding the decoded image of P8 by intra-frame encoding is
Combine immediately before the data after P11 extracted earlier and
OP♯ (r ′). And G in video stream 1
Streams before OP # (1), GOP # (1 '), G
OP {(r ') and GOP in video stream 2}
The streams after (r) are combined to obtain the edit stream 3.

FIG. 4 shows a specific example of an MPEG stream editing apparatus when performing the editing shown in FIG. In FIG. 4, A1 is a storage medium on which video stream 1 is recorded. From this medium A1, G in video stream 1
The image data preceding the OP # (1) and the data 107 of the GOP # (1 ') cut out from the GOP # (1) are read and written to the storage medium A3. Next, the storage medium A2
Is read out from the video data 101 of GOP # (r) in the video stream 2 recorded in the
1 is input. The decoder 41 decodes I2, P5, and P8 in order to obtain a decoded image 102 of P8. The decoded image 102 is intra-frame encoded by an encoder (encoder) 42, and an I picture signal 104 is obtained. Here, since the encoder 42 only performs intra-frame encoding, the vector search unit and the reference image memory as in the conventional example shown in FIG. 11 are unnecessary, and the DCT converter 421 and the VLC converter 422 are not necessary. It is composed of The I picture signal 104 obtained by the re-encoding is written to the storage medium A3.

Subsequently, the partial data after P11 in the GOP # (r) in the video stream 2 and the image data 106 in the part after the GOP # (r) in the video stream 2 are read out from the storage medium A3, Is written to. As described above, the edit stream 3 is obtained on the storage medium A3.

In the examples of FIGS. 3 and 4, the MPEG2 video stream can be edited in units of GOP or less, specifically, in units of intervals between I pictures or P pictures. In a normal MPEG2 video stream, the interval between I-pictures or P-pictures is about three video frames, so the unit of editing is about three video frames.

However, there may be a case where editing in a finer unit, that is, a video frame unit is required, for example, for the purpose of clearly connecting audios to be reproduced simultaneously. In consideration of such a case, the present invention provides a method of adjusting the reproduction time length of the edited portion in video frame units by adding picture data to the edit stream 3 shown in FIG.

In the embodiment shown in FIG. 5, coded image data is added in units of pictures to GOP # (1 ') data extracted from video stream 1. If large data is added here, overflow or underflow may occur in the VBV buffer in the edited video stream. For this reason, in this embodiment, an encoded image of a fixed pattern that does not increase the data amount and does not require encoding processing is added.

In the example of FIG. 5A, P picture data P9 without a vector and without a difference component is added to the end of GOP (1 '). The reproduced image of P9 is exactly the same as P8, and the code amount is extremely small. In the example of FIG. 5B, the I picture data I of a fixed pattern such as a black image
9 is added at the end of GOP $ (1 '). In this embodiment, I9 is included in GOP # (1 '), but this I picture may be defined as a GOP different from GOP # (1').

In any of the embodiments, the added picture is GOP # (1 ') extracted from the video stream 1.
Will be played immediately after. Further, in each embodiment, the added image data is only one picture, but data of a plurality of pictures can be added.

In the embodiment shown in FIG. 6, encoded image data is inserted in units of pictures into GOP♯ (r ′) data generated from the video stream 2. Here also, in the present invention, an encoded image of a fixed pattern that does not require a large amount of data and does not require encoding processing is added. In the embodiment of FIG. 6, backward prediction is performed without a vector and without a difference component (that is, inter-frame encoding that refers only to I8).
The B picture data B7 using only GOP @ (r ')
Is inserted immediately after the data of I8. The reproduced image of B7 is exactly the same as I8, and the code amount is extremely small. The added picture is reproduced immediately before I8. In this example, data of a plurality of pictures can be added.

As described above, according to the embodiment shown in FIGS. 3 and 4, the inter-frame coded video signal can be edited in units of GOP or less, and generated at the time of re-encoding. Since the pictures are only intra-frame coded pictures, it is possible to prevent an increase in the scale of a hardware device and an increase in software processing time in the re-encoding process. In addition, it is possible to prevent quality degradation of the re-encoded video signal due to the spread of noise on the reference image caused by inter-frame encoding.

According to the embodiment shown in FIGS. 5 and 6, FIG.
Since an image can be added in video frame units to the editing result of the embodiment in FIG. 4, as a result, editing in video frame units required when editing a video stream accompanied by an audio signal is performed. Can be realized.

[0065]

As described above, according to the present invention, an inter-frame coded video signal can be edited in a unit of GOP or less. It is possible to prevent an increase in time, prevent a deterioration in the quality of a video signal after re-encoding due to the influence of noise on a reference image caused by inter-frame encoding, and further edit a video stream with an audio signal. It is possible to provide a video stream editing method and apparatus capable of realizing editing in video frame units required in some cases.

[Brief description of the drawings]

FIG. 1 is a flowchart showing an embodiment of a video stream editing method according to the present invention.

FIG. 2 is a block diagram showing an embodiment of a video stream editing apparatus according to the present invention.

FIG. 3 is a timing chart showing an embodiment of an editing process according to the present invention.

FIG. 4 is a block diagram showing a configuration example for realizing the editing process of FIG. 3;

FIG. 5 is a timing chart showing another embodiment of the editing process according to the present invention.

FIG. 6 is a timing chart showing another embodiment of the editing processing according to the present invention.

FIG. 7 is a view for explaining an MPEG2 encoding method.

FIG. 8 is a view for explaining an MPEG2 encoding method.

FIG. 9 is a diagram showing a conventional video stream editing example.

FIG. 10 is a diagram showing another conventional example.

FIG. 11 is a diagram showing a configuration example for realizing the editing process of FIG. 10;

[Description of Code] 11, 12, 17: Video stream storage medium 13: Decoder (decoder) 14: Encoder (encoder) 15: DCT converter 16: VLC converter 18: Data recording device 19: Vector search unit 20 Reference image memory 21 IDCT converter A1, A2, A3 Video stream storage medium 31 First cutout processing unit 32 Second cutout processing unit 33 Picture type determination unit 34 Stream connection processing unit 41 Decoder (decoder) 42 ... Encoder (encoder) 421 ... DCT converter 422 ... VLC converter

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5C053 FA14 GA11 GB08 GB22 GB26 GB30 GB38 5C059 KK12 KK36 KK41 MA00 MA04 MA05 MA23 PP05 PP06 PP07 RB15 RC07 SS11 TA03 TA18 TA25 TB03 TC24 5J064 AA01 BA16 BB03 BC01 BC02 BD03

Claims (10)

[Claims] 1. A video stream editing method for obtaining a continuously reproducible video stream by concatenating a part or all of a first video stream and a part or all of a second video stream which have been subjected to inter-frame prediction coding. A first procedure of cutting out a first partial video stream from the first video stream up to immediately before a picture coded by intra-frame predictive coding or unidirectional inter-frame predictive coding; A second sub-video stream for extracting a second partial video stream after the picture coded by intra-frame coding or unidirectional inter-frame prediction from the video stream;
And a third step of determining whether the picture displayed immediately before the second partial stream cut out from the second video stream in the second step is an intra-frame coded picture or not. And if it is determined that the picture displayed immediately before the second partial stream cut out from the second video stream in the third procedure is an intra-frame coded picture, This picture is defined as a first intra-coded picture, and if it is determined that the picture is not an intra-coded picture, unidirectional inter-frame prediction from the intra-coded picture immediately before the picture to the picture is performed. After obtaining a decoded image of the picture by decoding the coded picture in order, the decoded picture is re-encoded by intra-frame coding. A fourth procedure of outputting a picture obtained by this re-encoding as a first intra-coded picture; a first partial stream obtained by the first procedure; and a fourth stream obtained by the second procedure. A fifth procedure for inserting the first intra-coded picture obtained in the fourth procedure between the second partial stream and the second partial stream. 2. The video stream editing method according to claim 1, wherein the fifth step includes the step of adding at least one piece of predetermined picture data immediately after the first partial video stream, and A video stream editing method, characterized in that a stream is generated by sequentially connecting an intra-coded picture and a second partial video stream. 3. The video stream editing method according to claim 2, wherein the picture data to be added immediately after the first partial video stream includes the last display picture of the first partial video stream as a reference picture. A video stream editing method, characterized in that the pictures are inter-frame predictively coded pictures so as to display the same image. 4. The video stream editing method according to claim 1, wherein the fifth step comprises the step of:
And adding the at least one picture of a predetermined bidirectional prediction type following the first intra-coded picture, and concatenating the second partial video stream. Method. 5. The video stream editing method according to claim 4, wherein the bidirectional prediction type picture data to be inserted after the first intra-frame coded picture is the first intra-frame coded picture. A video stream editing method, characterized in that it is a picture that has been subjected to inter-frame prediction coding so as to display the same picture as the picture based on only the backward prediction to be referred to. 6. A video stream editing apparatus which obtains a continuously playable video stream by concatenating part or all of a first video stream and a part or all of a second video stream which have been subjected to inter-frame prediction coding. A first cutout processing unit for cutting out a first partial video stream from the first video stream up to immediately before a picture coded by intra-frame predictive coding or unidirectional inter-frame predictive coding; From the second video stream, a second partial video stream after a picture encoded by intra-frame encoding or unidirectional inter-frame prediction.
And whether or not the picture displayed immediately before the second partial stream cut out from the second video stream by the second cutout processing means is an intra-frame coded picture. A picture type determining means for determining, and a picture displayed immediately before the second partial stream cut out from the second video stream is determined to be an intra-frame coded picture by the picture type determining means. In this case, this picture is set as the first intra-coded picture, and if it is determined that the picture is not an intra-coded picture, one picture from the intra-coded picture immediately before the picture to the picture is determined. After obtaining a decoded image of the picture by sequentially decoding the directional inter-frame prediction coded picture, First intra-frame coded picture generating means for re-encoding the picture by intra-frame encoding, and outputting the picture obtained by the re-encoding as a first intra-frame coded picture; The first partial stream obtained by the processing means and the second partial stream obtained by the second cutout processing means.
Video stream editing means for inserting a first intra-coded picture obtained by the first intra-coded picture generation means between the partial stream of apparatus. 7. The video stream editing apparatus according to claim 6, wherein the stream connection processing unit adds at least one piece of predetermined picture data immediately after the first partial video stream, and then outputs the first partial video stream. An editing apparatus for a video stream, wherein a stream is generated by sequentially connecting an intra-frame coded picture and a second partial video stream. 8. The video stream editing apparatus according to claim 7, wherein the picture data to be added immediately after the first partial video stream includes the last display picture of the first partial video stream as a reference picture. A video stream editing apparatus, characterized in that the pictures are inter-frame predictively coded pictures that display the same image. 9. The video stream editing apparatus according to claim 6, wherein said stream concatenation processing means comprises, following said first partial video stream and said first intra-coded picture, at least one A video stream editing apparatus, wherein after adding a picture of a predetermined bidirectional prediction type, the second partial video stream is connected. 10. The video stream editing apparatus according to claim 9, wherein the bidirectional prediction type picture data to be inserted after the first intra-coded picture is the first intra-coded picture. A video stream editing apparatus, which is a picture that has been subjected to inter-frame prediction coding so as to display the same image as the picture based on only the backward prediction to be referred to.