JP3505349B2 - Video compression encoding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [0001] BACKGROUND OF THE INVENTION 1. Field of the Invention
Converts a field image from a 24 Hz progressive scanning video signal.
Original of NTSC system generated by repeating periodically
Video compression encoding apparatus for encoding an image signal as input
Especially for multi-angle original image signals.
Video compression coding that can prevent the generation of a long code amount
Equipment related. [0002] 2. Description of the Related Art In recent years, images are compressed by digital signals.
Various technologies have been developed. Of these, computers, radio
Applications used in the three fields of
The method considered to satisfy all is called MPEG2.
This is a moving image compression encoding method. This MPEG2 way
For the moving image compression coding by the expression, for example, “IS
O / IEC 13818-2 Draft International St
anderd ". This method smells
The DCT method of 8 × 8 pixels in one frame
GOP in addition to compression encoding (intra-frame compression encoding)
Multiple image frames called (Group of Pictures)
Is provided, and the compression encoding is performed.
I'm trying to do it. [0003] A moving picture compression code of the MPEG2 system is used.
DVD (Digital Video Disk)
There is an encoder. This DVD encoder uses M
PEG2 format is adopted, and Dolby Digital format or Li
The near PCM method is used, and compression encoding is performed using those methods.
Video and audio bitstreams into MPEG2
Format to the DVD standard based on the system layer.
Device. In addition, DVDs are multi-story
About the angles, the menus that accompany them, etc.
Supports and has interactive features. In this case, according to the DVD standard, multi-angle
Video bitstreams for
includedGOPThe number of pictures in the
The number of frames needs to be constant. Also, these
Logical format of bit stream to DVD standard signal
The same time code to interleave
That the GOP code including the code
Is desired. Therefore, the original image signal for multi-angle
Usually performs compression encoding at a fixed bit rate.
I do. [0005] In addition, the DVD encoder is a multi-angular encoder.
The original image signal for the movie is a movie material 映像 24 frames per second
To 30 frames per second (29.
97 frames) video signal
Performs compression encoding after detecting 3: 2 pulldown
Run. Here, such 3: 2 pull-down detection is performed by:
Originally, a field image is repeated for an image with little motion.
If it is not returned, it will be repeatedly judged as an image
There is a possibility. Therefore, using this detection method,
Encodes multiple original image signals for tilt
And the number of pictures in the GOP differs due to this false detection.
As a result, the DVD standard is not satisfied. This
To solve this problem, conventional DVD encoders use
When compressing and encoding the original image signal for multi-angle:
2. Encoding was performed without detecting the pull-down. This is a mark
In each GOP of the encoded video bitstream
This is a process for keeping the number of keys to be constant.
2 In encoding without pull-down detection, field repetition
Redundant codes are used because the returned image is re-encoded.
However, there is a problem that the amount is generated and the image quality is deteriorated. [0006] The above-mentioned conventional DVD
In the encoder, false detection of 3: 2 pull-down detection
As a result, the number of pictures in the GOP differs,
Multi-angle to prevent losing the case
3: 2 pulldown when compressing and encoding the original image signal for
Encoding without detection, but in this case the field
Redundantly re-encodes the image.
There is a problem that a code amount is generated and image quality is deteriorated. [0007] The present invention eliminates the above-mentioned problems, and provides a multi
3: 2 pulldown detection of original image signal for angle and compression
Prevent the number of pictures in a GOP from differing when encoding
It is an object of the present invention to provide a moving image compression encoding apparatus capable of performing the above. [0008] SUMMARY OF THE INVENTION Moving picture compression according to the present invention
The encoding device is a progressive scanning video with a frame frequency of 24 Hz.
Image signalOf each field of two fields of the same image
Return image or repeated image of 3 fields of the same image
Generated by convertingNTSC original image signal
A moving image compression encoding apparatus to be input, and before inputting
A repetitive image of the original image signal of the NTSC system is detected,
Perform 3: 2 pulldown on detected repeated images
In addition, by performing field-frame conversion,
3: 2 pulldown detection that creates and outputs image signals
And the image signal output by the 3: 2 pull-down detection circuit.
A compression-encoding circuit for compression-encoding;
Create pull-down information that is the detection result of the
3: 2 pull-down information creating and saving means for saving;
3: 2 pull, which is the detection result of the 3: 2 pull-down detection circuit
When there are a plurality of down information, a plurality of 3: 2
During the period when the pull-down information is common, the 3: 2 pull-down
Down-down information in the 3: 2 pull-down information creation storage means
In the period when multiple 3: 2 pulldown information are different,
Set the 3: 2 pull-down phase of the 3: 2 pull-down information to 2
Change to pull-down phase and save 3: 2 pull-down information
3: 2 pull-down information comparison and storage means
And 3: 3 from the 3: 2 pull-down information comparison storage means.
2 According to the pull-down information,
The compression encoding circuit performs 3: 2 pull-down of an image signal
3: 2 pull-down means for supplying to the
is there. [0009] BRIEF DESCRIPTION OF THE DRAWINGS FIG.
It will be described with reference to FIG. FIG. 1 is a moving picture compression encoding apparatus according to the present invention.
It is a block diagram showing an embodiment. FIG. 1 shows a moving image according to an embodiment of the present invention.
The image compression encoding apparatus has a frame frequency of 24 Hz sequentially.
Periodically repeat field image from scanned video signal
The original image signal a1 of the NTSC system generated by the
I have strength. Reference numeral 11 denotes a multi-angle NTSC system.
Input terminal to which an original image signal a1 is input.
The original image signal a1 guided to the input terminal 11 is 3: 2 pull
It is supplied to the down detection circuit 12. The 3: 2 pull-down detection circuit 12 has an input terminal
The original image signal a1 guided to the
Based on 3: 2 pulldown detection mode selection signal b1
Perform 3: 2 pulldown and perform field-frame conversion
The image signal c1 is generated and supplied to the compression encoding circuit 14.
You. Here, the 3: 2 puller from the input terminal 13
The power detection mode selection signal b1 is in the automatic detection mode.
Is a binary signal that is true for a set of multi-angle
Original image signal for the first multi-angle
True when compression encoding a signal
False, the original image signal for non-multi-angle
True if The 3: 2 pull-down detection circuit 12
When the 3: 2 pull-down detection mode selection signal b1 indicates true,
In this case, 3: 2 pulldown is detected, that is, the original
Detects a repetitive image of the image number, and based on this detection result
Field-to-frame conversion with 3: 2 pulldown
To generate the image signal c1 and supply it to the compression encoding circuit 14.
Time code information when pulling down
And the time code image is a field repeat image
Whether the 3: 2 pull-down phase information indicates at least
Create 3: 2 pull-down information d1 including the recording means 1
Record in 5. The 3: 2 pull-down detection circuit 12
If the 2 pull-down detection mode selection signal b1 indicates false,
In this case, the 3: 2 pull-down information of the recording means 15 previously generated
Determine the phase of 3: 2 pulldown using d1, and
Perform 3: 2 pulldown on image signal a1
An image signal c1 is created by performing frame conversion to generate a compression code;
To the conversion circuit 14. The compression / encoding circuit 14 has a 3: 2 pull-down
The image signal c1 from the detection circuit 12 is compression-encoded and
A bit stream signal e1 is created and output, and a recording medium
Then, a video bitstream file is generated at step S16. With such a configuration, 3: 2 pull-down
The detection circuit 12 receives the NTSC original image signal.
No. a1 repeated image is detected, and the detected repeated image
3: 2 pull-down for
-Create an image signal c1 by performing frame conversion
3: 2 pull-down detection circuit that outputs
The conversion circuit 14 outputs the 3: 2 pull-down detection circuit 12.
A compression encoding circuit for compressing and encoding the input image signal c1;
The 3: 2 pull-down detection circuit 12 and the recording means 15
Is a detection result of the 3: 2 pull-down detection circuit 12.
3: 2 Create and save 3: 2 pulldown information d1
It becomes a pull-down information creation storage means. Also 3: 2 pull
The down detection circuit 12 generates the 3: 2 pull-down information.
Other input according to 3: 2 pulldown information from storage means
3: 2 puller of the original image signal a1 of the NTSC system
3: 2 program to be supplied to the compression / encoding circuit 14
It is also a means of down. The operation of the embodiment of the present invention will be described below.
explain. First original image signal a1 for multi-angle
Is input from the input terminal 11, 3: 2 pull down
Detection mode signal b1 becomes true, and 3: 2 pull-down detection
The output circuit 12 provides a 3: 2 pulldown for the original image signal a1.
And a 3: 2 pull-down and
And field-to-frame conversion to create an image signal c1
And supplies it to the compression encoding circuit 14,
Field information and its time code image
3: 2 pulldown phase information indicating whether the image is a return image
Create and record 3: 2 pulldown information d1 including at least
Recording means 15. The compression encoding circuit 14 is supplied
Compression and coding of the video signal c1
No. e1 is created and output, and the video bit
Generate a trim file. The next original image signal a1 for multi-angle is
3: 2 pulldown when input from input terminal 11
The detection mode signal b1 becomes false, and 3: 2 pulldown detection is performed.
The circuit 12 stores in the recording means 15 the 3: 2 puller generated earlier.
3: 2 pull-down phase is determined using the download information d1
Then, based on this determination, the 3: 2 pull-down of the original image signal a1 is performed.
Image signal c1
And supplies it to the compression encoding circuit 14. Compression coding
The circuit 14 converts the supplied image signal c1 to a fixed bit rate.
To create a video bit stream signal e1
Video bit stream file on the recording medium 16.
Generate files. The original image signal for the next multi-angle
Similarly, when the signal a1 is compression-encoded,
Pull down using 3: 2 pull down information d1
Perform compression encoding. According to such an embodiment of the present invention,
Same for a plurality of original image signals a1 for multi-angle
3: 2 pulldown using 3: 2 pulldown information d1
The original image signal for multi-angle
GOP when detecting 3: 2 pulldown and compressing and encoding
｛(Group of Pictures), composed of multiple image frames
Prevents the number of pictures in the generated coding unit 異 な る from differing
Yes, the frame frequency is 24Hz
Next scan signal is converted to NTSC television signal
When the converted signal is used as the original image signal for multi-angle
Even in this case, it is possible to execute 3: 2 pulldown and encode
As a result, the amount of redundant code can be reduced and the image quality can be improved.
it can. FIG. 2 is an original image signal for multi-angle of FIG.
FIG. 3 is an explanatory diagram for explaining compression encoding. In FIG. 2, four of the multi-angle periods
Of the original image signals 91, 92, 93 and 94
It is. The original image signal 90 before and after the multi-angle period,
At 95, the signal is a single system. Original image signals 91 and 9
2, 93 and 94 are compression-encoded by the MPEG2 system,
Video bit streams 96, 97, 98, 99 respectively
Is converted to In this case, according to the DVD standard, each generated
Video bit streams 96, 97, 98, 99 are the same
Includes timecodeGOPNumber of pictures in
And the number of display frames must be constant.
You. These bit streams 96, 97, 9
When logically formatting 8,99 into a DVD standard signal
Contains the same time code to interleave
It is hoped that the code amount of the GOP
You. Therefore, the original image signals 91 and 9 for multi-angle
2, 93 and 94 are fixed as in the embodiment of the invention of FIG.
Perform compression encoding at a constant bit rate. The 3: 2 pull-down detection circuit 12
Use FIGS. 3 and 4 for the outline of 3: 2 pulldown detection
Will be described. FIG. 3 shows a 24 Hz progressive scan of a movie material.
How to convert signals to NTSC signalsShowFIG.
FIG. 3A shows an original signal of 24 Hz, and FIG.
Indicates a signal after NTSC conversion. In FIG. 3, a movie material of 24 Hz
The progressive scanning signal is converted to a frame frequency of 29.97 Hz NT.
When converting to an SC signal, the 24H signal shown in FIG.
3 (b) from the first frame of the original signal of zShown inNTS
Create signals of the first to third fields of the C-converted signal
The second frame of the original signal of 24 Hz shown in FIG.
Figure 3 (b)Shown inFourth of NTSC converted signal
And the signal of the fifth field are created and shown in FIG.
From the third frame of the original signal of 24 Hz, FIG.Shown in
NTSC-converted signal6th to 8thField trust
No. 4 of the original signal of 24 Hz shown in FIG.
Fig. 3 (b) from the frameShown inNTSC converted signal
ofNinth and tenthCreate a signal for the field and follow the diagram below
The first to fourth frames of the original signal shown in FIG.
The operation up to is repeated. In this case, the first field and the third field
Field becomes the same image, the sixth field and the eighth field
Are the same image. This NTSC signal periodically has the same field.
It is made up of repeated images. Follow
Therefore, the signal as shown in FIG.
This video file is input to the
Field repetition image is detected.
Performs redundant encoding by performing encoding only once.
The generation of an amount can be avoided. And that fee
3: 2 pull down detection
I'm calling FIG. 4 shows an example of a 3: 2 pull-down detection method.
It is explanatory drawing explaining the outline of. In FIG. 4, each image of the k-th field image
The prime values are pa1, pa2,... Pa (n-1), pan
And the pixel value of the (k + 2) th field image is pb
1, pb2,... Pbn. And this (k +
2) The field image is a repeated image of the k-th field image
First, to indicate whether an image is
Absolute value of the difference | pa1-pb1 |, | pa2-pb2 |,
... | pan-pbn | Then, those values
Less than a certain levelToCalculate the number of pixels. Sand
That is, the number of pixels is determined by the (k + 2) -th field image
The degree of similarity to the field image is shown. Soshi
If the pixel value is greater than the threshold level
Means that the (k + 2) th field image is the kth field image
Is determined to be a repeated image. FIG. 5 is a moving picture compression encoding apparatus according to the present invention.
It is a block diagram which shows 2nd Embodiment of this. In FIG. 5, reference numeral 21 denotes an NTSC system.
An input terminal to which an original image signal a2 for multi-angle is input
And an original image signal a2 guided to the input terminal 21.
Is supplied to the 3: 2 pull-down detection circuit 22. The 3: 2 pull-down detection circuit 22 has an input terminal
The original image signal a2 guided to the child 21
Based on 3: 2 pulldown detection mode selection signal b2
Perform 3: 2 pulldown and perform field-frame conversion
The image signal c2 is generated and supplied to the compression encoding circuit 24.
You. Here, the 3: 2 puller from the input terminal 23
Detection mode selection signal b2 is in the automatic detection mode.
Is a binary signal that is true for a set of multi-angle
Original image signal for the first multi-angle
True when compression encoding a signal
False, the original image signal for non-multi-angle
True if The 3: 2 pull-down detection circuit 22
When the 3: 2 pull-down detection mode selection signal b2 indicates true,
In this case, 3: 2 pulldown is detected, and based on this detection result,
Perform 3: 2 pulldown and perform field-frame conversion
The image signal c2 is generated and supplied to the compression encoding circuit 24.
You. The 3: 2 pull-down detection circuit 22 outputs a 3: 2 pull-down signal.
If the drop-down detection mode selection signal b2 indicates false,
The 3: 2 pull-down information d2 of the recording means 25 generated in
Determine the phase of 3: 2 pulldown using 3: 2 pulldown
Down, field-frame conversion, image signal
c2 is created and supplied to the compression encoding circuit 24. The compression / encoding circuit 24 has a 3: 2 pull-down
The image signal c2 from the detection circuit 22 is compression-encoded and
A bit stream signal e2 is created and output, and a recording medium
At 26, a video bit stream file is generated. The 3: 2 pull-down information extraction circuit 27
First multi-angle video recorded on recording medium 26
Time code information and 3: 2
Generate 3: 2 pulldown information d2 including pulldown phase information
And record it in the recording means 25. With such a configuration, the recording means 25 and
The 3: 2 pull-down information extraction circuit 27 is a compression encoding circuit
24 from the video output result compressed and coded in
Information and 3: 2 pulldown phase information.
Existing 3: 2 pull-down information extraction storage means
The 2 pull-down detection circuit 22 detects the 3: 2 pull-down information.
Time code information and 3: 2
The NTSC method for inputting another according to the down-down phase information
3: 2 pulldown of the original image signal of the formula
3: 2 pull-down means to supply to the conversion circuit 24
I have. FIG. 6 shows the operation of the embodiment of the present invention.
This will be described below with reference to FIG. FIG. 6 shows the 3: 2 pull-down information extraction time of FIG.
9 is a flowchart showing an algorithm of the operation of the road 27.
You. As shown in FIG. 6, 3: 2 pull-down information
The extraction circuit 27 determines in step S1 that the video bit stream
The detection determination of the GOP header in the stream is executed. here
GOP is a coding unit used in the MPEG2 system or the like.
In the GOP header, which is the header,
The first picture displayed in the encoded picture of
Key time code is described. 3: 2 pulldown
The information extraction circuit 27 detects no GOP header
In this case, the determination in step S1 is no and the process ends.
However, if a GOP header is detected, the determination in step S1 is made.
The determination is yes, and the process proceeds to step S2. In step S2, 3: 2 pull-down
The information extraction circuit 27 obtains the time code in the GOP.
Then, the process proceeds to step S3. In step S3, 3: 2 pull-down
The information extraction circuit 27 detects the picture header in the GOP.
I do. If a picture header is detected, step
If the determination in S3 is yes, the process proceeds to step S4
I do. The header for each picture contains the picture
GOPThe order that indicates the position in the
ing. Others include the field picture of the picture.
Whether the image is a repetitive image,
3: 2 pull-down phase information such as whether to start with
include. The 3: 2 pull-down information extraction circuit 27
In step S4, when these pieces of information are obtained,
Then, the process returns to the detection of the picture header in S3. The above steps
The operations of S3 and S4 are detected when the picture header is detected in the GOP.
Repeat until no more. Picture header in GOP
If the detection has stopped, the determination in step S3 is NO.
Then, the process proceeds to step S5. In step S5, 3: 2 pull-down
The information extraction circuit 27 determines whether the picture header is
Number, 3: 2 phase information and step S
1 GO from the time code obtained from the GOP header in 2
Calculate the time code of the picture for P and at least
Calculated time code and 3: 2 pulldown phase information
Is generated as 3: 2 pull-down information, and
Record and return to GOP header detection in step S1 again
You. Then, the GOP header is detected from the video bit stream.
The process ends when no more can be output. This allows
Extract the time code and 3: 2 phase information of the picture
3: 2 pull-down information shown in Fig. 5 can be generated.
You. Next, the operation of the entire moving picture compression encoding apparatus will be described.
explain. First original image signal a2 for multi-angle
Is input from the input terminal 21, 3: 2 pull down
Detection mode signal b2 becomes true, and 3: 2 pull-down detection is performed.
The output circuit 22 provides a 3: 2 pulldown for the original image signal a2.
And a 3: 2 pulldown is performed based on this detection result.
And performs field-frame conversion to produce an image signal c2.
The data is supplied to the compression encoding circuit 24, where the encoded data is compressed and encoded.
A video bit stream file is generated on the recording medium 26.
You. The 3: 2 pull-down information extraction circuit 27 stores the recording medium 2
6 video bits for the first multi-angle recorded in
Generate 3: 2 pulldown information d2 from the stream and record it
Recorded in the means 25. The next original image signal a2 for multi-angle is
3: 2 pulldown when input from input terminal 21
The detection mode signal b2 becomes false, and 3: 2 pulldown detection is performed.
The circuit 22 includes the 3: 2 puller generated before the recording means 25.
3: 2 pull-down phase is determined using the download information d2.
Then, based on this determination, 3: 2 for the original image signal a2.
Perform pull-down and field-frame conversion to image
A signal c2 is created and supplied to the compression encoding circuit 24 for compression.
Encoded and stored in the recording medium 26 as a video bit stream file.
Generate files. The original image signal for the next multi-angle
Similarly, when compressing and encoding signals, 3: 2 pulldown information
Information to perform compression encoding. 3: 2 pull made in advance as described above
Determining the phase of 3: 2 pulldown with down information
From the original image signal for multi-angle 3: 2 pulldown
When the number of pictures in a GOP is
Because it can prevent the difference, frame like movie material
The signal of the progressive scanning method with the frequency of 24 Hz is converted to the NTSC method.
For multi-angle television signals
3: 2 pulldown is executed even when the original image signal is
Encoding, reducing the amount of redundant code and improving image quality
Can be done. FIG. 7 is a moving picture compression encoding apparatus according to the present invention.
FIG. 10 is a block diagram showing a third embodiment of the present invention,
Edit is included in one of the original image signals for angle 3: 2
This is for when the pull-down phase is out of order.
You. In FIG. 7, reference numeral 31 denotes an NTSC system.
An input terminal to which an original image signal a3 for multi-angle is input
And an original image signal a3 guided to the input terminal 31.
Is supplied to the 3: 2 pull-down detection circuit 32. This place
Original image for multi-angle guided to input terminal 31
The signal a3 is a signal that has not been edited at first, and
Will be edited. The 3: 2 pull-down detection circuit 32 has an input terminal
The original image signal a3 guided to the child 31
Based on 3: 2 pulldown detection mode selection signal b3
Perform 3: 2 pulldown and perform field-frame conversion
The image signal c3 is generated and supplied to the compression encoding circuit 34.
You. Here, the 3: 2 puller from the input terminal 33
Signal detection mode selection signal b3 is in the automatic detection mode.
Is a binary signal that is true for a set of multi-angle
Original image signal for the first multi-angle
No. (no editing) and original image signal for second multi-angle
True when compression-encoding a code (with edits)
False when compression encoding from non-multi-angle
Is true in the case of the original image signal. The 3: 2 pull-down detection circuit 32 outputs the 3: 2
When the pull-down detection mode selection signal b3 indicates true,
3: 2 pull-down is detected, and based on the detection result, 3: 2 pull-down is detected.
2 Perform pull-down, perform field-frame conversion, and
When the image signal c3 is created and supplied to the compression encoding circuit 34,
In both cases, the time code information and its
Whether the timecode image is a field repeat image
Contains at least 3: 2 pull-down phase information
Create 3: 2 pulldown information d3, original image without editing
The signal is recorded in the first recording area 41 of the recording means 35.
For the original image signal recorded and edited, the recording means 35
Recording is performed in the second recording area 42. 3: 2 pulldown detection
The circuit 32 has a 3: 2 pull-down detection mode selection signal b3
Indicates false, 3: 2 of the recording means 43 generated earlier.
3: 2 pulldown phase using pulldown information d3
And perform 3: 2 pulldown to perform field-frame
The image signal c3 is generated by performing the
4 The compression encoding circuit 34 has a 3: 2 pull-down
The image signal c3 from the detection circuit 32 is compression-encoded and
Outputs the bit stream signal e3 and the original image without editing
Recorded for signal and original image signal with edit except the first
The bit stream file is stored in the first recording area 44 of the medium 36.
The original image signal with the first edit
A bit stream is stored in the second recording area 45 of the recording medium 36.
Generate a file. 3: 2 pulldown information comparison and generation circuit 37
Are the first and second recording areas 41 and 42 of the recording means 35.
When 3: 2 pulldown information exists in both
Of the first and second recording areas 41 and 42.
3: 2 pulldown as it is during the period when the information is shared
Information is stored in the recording means 43, and the first and second recording
The 3: 2 pulldown information of the areas 41 and 42 is not common.
3: 2 pulldown of 3: 2 pulldown information
The recording phase to the pull-down phase of 2
Let me save. First and second recording areas 4 of recording medium 36
The bit stream files recorded in 4, 45 are
Decoded and expanded by a decoding circuit (not shown)
3: 2 pulldown detection again as single angle original image signal a3
The circuit 32 outputs the 3: 2 pull-down information d3 of the recording means 43.
Using 3: 2 pulldown and field-frame conversion
Is performed, and the image signal c3 is sent to the compression encoding circuit 34.
Supplied and recorded in the first recording area 44 of the recording medium 36.
It is. With such a configuration, 3: 2 pull-down
The detection circuit 32 and the recording means 35 and 43 are provided with a 3: 2 puller.
The pull-down information, which is the detection result of the
Create and save 3: 2 pulldown information creation and storage means
3: 2 pull-down information comparison and generation circuit 37
3: 2 which is the detection result of the 3: 2 pull-down detection circuit 32
When there are multiple pull-down information,
As long as 3: 2 pulldown information is common
Create 3: 2 pulldown information and save 3: 2 pulldown information
Means, multiple 3: 2 pulldown information is different
In the period, 3: 2 pulldown rank of 3: 2 pulldown information
Change the phase to 2 pull-down phase and select 3: 2 pull-down information
3: 2 pull-down information comparison to be stored in report creation storage means
The 3: 2 pull-down detection circuit 32 serves as storage means.
3: 2 pulldown from 3: 2 pulldown information comparison storage means
According to the down information, the original image signal of the NTSC system is renewed.
3: 2 pull-down of the signal to the compression / encoding circuit 34
3: 2 pull-down means for supplying. The 3: 2 pull-down information comparison and generation circuit 37
The determination operation will be described with reference to FIG. FIG. 8 shows a 3: 2 pull-down information comparison / generation circuit.
FIG. 8 is an explanatory diagram showing a pull-down phase for making a determination, and FIG.
(A) shows a multi-angle original image signal a3 without editing.
FIG. 8B shows a multi-angle original image signal with editing.
FIG. 8C shows a pull-down phase of 2
I have. First, a multi-angle original image signal without editing
The signal a3 has a phase ratio of 3: 2 as shown in FIG.
They are lined up regularly. On the other hand, multi
The angle original image signal a3 is, as shown in FIG.
Edited in two places, marked A1 and B1, as shown in the figure, for example.
It is assumed that the 3: 2 phase is discontinuous as described above. This
When the phase to be generated, the phase of 3: 2 starts to be different
8c, the two phases shown in FIG.
If there is a place where the phase matches again, up to that point, the phase of 2
To the end. By the above operation, 3: 2
3: 2 pulldown can be performed during the period when
Therefore, the conventional compression code that does not perform 3: 2 pulldown at all
The image quality can be improved as compared with the decoding circuit. FIG. 9 shows a 3: 2 pull-down information comparison and generation circuit.
37 is a flowchart showing a specific algorithm example of 37.
is there. In FIG. 9, 3: 2 pull-down information comparison
The generation circuit 37 determines in step S11 that the recording unit 3
5 is 3: 2 in both the first and second recording areas 41 and 42.
If there is pull-down information, the
It is determined whether or not the imcode has been acquired. Both pics
If the time code of the cha could not be obtained,
The determination in S11 is no, and all the processing is completed.
If the time code of the picture has been acquired, step S1
The determination of 1 is affirmative and the process proceeds to step S12.
You. In step S12, 3: 2 pull down
The information comparing / generating circuit 37 includes first and second
3: 2 pulldown recorded in 2 recording areas 41 and 42
Get the timecode of both pictures in the information,
It is determined whether these time codes match. both
If the time codes of the pictures do not match,
The determination in step S12 becomes no, and the processing in step S13 is performed.
If the transition is made and the timecodes of both pictures match,
In this case, the determination in step S12 is YES, and step S1
The processing moves to the processing of step 4. In step S14, 3: 2 pull down
The circuit information comparison and generation circuit 37 generates a picture of the same time code.
Field to force 3: 2 phase match
2 in a 3: 2 phase to stop image repetition
Change the pattern to In some cases, 3: 2
To adjust the phase, the 3: 2 phase of the previous picture is set to 2
It is also necessary to change to And those timecos
Code and 3: 2 phase as 3: 2 pulldown information
Then, it records it in the recording means 43 and returns to the processing of step S11.
You. In step S14, 3: 2 pull down
The comparison information generation circuit 37 determines that the 3: 2 phase of the picture is
Check for a match. 3: 2 phase matches
In this case, the 3: 2 pull-down information comparison and generation circuit 37
Shift to the processing of step S15, and place 3: 2 in the picture
Generate phase and time code as 3: 2 pulldown information
Then, it records it in the recording means 43 and returns to the processing of step S11.
You. 3: 2 pulldown when 3: 2 phase does not match
The information comparison and generation circuit 37 proceeds to the process of step S16
Then, as in the case of time code mismatch, 3: 2
Match the phases to the pattern of 2: 3: 2
Information is generated and recorded in the recording means 43,
It returns to the process of 11. A 3: 2 pulldown information comparison student as described above
By using the generating circuit 37, the common 3: 2 phase
Turns are preserved, otherwise 3: 2 phase
By setting to 2 patterns, all multi-angle
It is possible to match the 3: 2 pulldown information of the original image signal.
it can. The operation of the embodiment of the present invention will be described below.
Will be described. First, the first edited multi-angle
The original image signal a3 which has not been input is input from the input terminal 31.
In this case, the 3: 2 pull-down detection mode selection signal b3 is set to true.
The 3: 2 pull-down detection circuit 32 outputs the original image signal a
3: 2 pulldown is detected for 3 and based on this detection result.
And perform 3: 2 pull-down to create an image signal c3.
First recording on the recording medium 26 via the compression encoding circuit 34
When a video bitstream file is generated in the area 44,
In both cases, the 3: 2 print is stored in the first recording area 41 of the recording means 35.
Record downtime information. Next, multi-angle editing is performed.
When the original image signal a3 is input from the input terminal 31,
The 3: 2 pull-down detection circuit 32 detects the original image signal a3.
Then, 3: 2 pull-down is detected, and based on the detection result,
A 3: 2 pulldown is performed to create an image signal c3, and a compression code
The second recording area 4 of the recording medium 26 via the encoding circuit 34
5 to generate a video bitstream file
In the second recording area 42 of the recording means 35, the 3: 2 pull
Record information. Thus, 3: 2 pulldown information comparison student
The forming circuit 37 is provided for the first and second recording areas 41 and 42.
As long as 3: 2 pulldown information is common
The 3: 2 pulldown information is stored in the recording unit 43, and the first
And 2: 2 pulldown information of the second recording areas 41 and 42
Are not common, the 3: 2 pulldown information
Change 3: 2 pulldown phase to 2 pulldown phase
To be stored in the recording means 43. Next, from the input terminal 31, a multi-angle
Again input the unedited original image signal a3 of
3: 2 pull-down detection mode selection signal b3 is false,
The 3: 2 pull-down detection circuit 32 outputs
Determine 3: 2 phase using 3: 2 pull-down information d3
Then, 3: 2 pulldown is performed on the original image signal a3 to
Create an image signal c3 and record it via the compression encoding circuit 34
The video bit stream is stored in the first recording area 44 of the medium 36.
Generate a file. Below, for the remaining multi-angle
Similar processing is applied to the unedited original image signal a3.
And the decoded original image read from the recording means 31
The same processing is performed on the signal a3 to perform recording.
Using the 3: 2 pull-down information d3 from the means 43, 3:
3: 2 pulldown operation with two phases determined and compression encoding
Operation for all multi-angle original image signals
And sequentially. With the above operation, the source for multi-angle
If any of the image signals a3 has been edited
Also during the period when the 3: 2 pull-down phase matches:
2 When performing compression encoding by performing pull-down,
Since the number of textures can be prevented from differing, redundant code
That period can be reduced and the image quality can be improved
it can. FIG. 10 shows a moving picture compression encoding apparatus according to the present invention.
FIG. 14 is a block diagram showing a fourth embodiment of the present invention,
Editing is included in one of the original image signals for the angle 3:
3: 2 pulldown when phase of 2 pulldown is disturbed
3: 2 pull-down information, which is the detection result of
It is generated from the stream and used. In FIG. 10, reference numeral 51 denotes an NTSC system.
Is the input to which the original image signal a4 for multi-angle is input
Terminal. The 3: 2 pull-down detection circuit 52 has an input terminal
The original image signal a4 guided to the
Based on 3: 2 pulldown detection mode selection signal b4
Perform 3: 2 pulldown and perform field-frame conversion
A video signal c4 is generated and supplied to the compression encoding circuit 54.
You. Here, the 3: 2 puller from the input terminal 53
The down detection mode selection signal b4 is the 3: 2 program shown in FIG.
This is the same as the drop-down detection mode selection signal b3.
The 3: 2 pull-down detection circuit 52 detects the 3: 2 pull-down.
When the output mode selection signal b4 indicates true, the original image signal a4
3: 2 pull-down is detected, and based on this detection result,
Field-to-frame conversion with 3: 2 pulldown
To generate an image signal c4 and supply it to the compression encoding circuit 54.
Pay. The 3: 2 pull-down detection circuit 52 outputs
When the 2 pull-down detection mode selection signal b4 indicates false
In this case, the 3: 2 pull-down information of the recording means 55 previously generated
Using d4, determine the phase of 3: 2 pulldown and 3:
2 Perform pull-down, perform field-frame conversion, and
The image signal c4 is created and supplied to the compression encoding circuit 54. The compression encoding circuit 54 has a 3: 2 pull-down
The image signal c4 from the detection circuit 52 is compression-encoded and
Outputs the bit stream signal e4 and the original image without editing
For original signals with editing except for the first
The bit stream file is stored in the first recording area 64 of the recording medium 56.
Generates a file and records the original image signal with editing
The bit stream file is stored in the second recording area 65 of the medium 56.
Generate a file. The 3: 2 pull-down information extraction circuit 57
The first mask recorded in the first recording area 64 of the recording medium 56
3: 2 pull from video bitstream for multi-angle
The down information f4 is generated, and the first recording area of the recording unit 55 is generated.
Recording in the area 61 and the second recording on the recording medium 56
Video for the second multi-angle recorded in area 65
Generates 3: 2 pulldown information f4 from bit stream
Then, the data is recorded in the first recording area 62 of the recording means 55. 3: 2 pulldown information comparison and generation circuit 58
Are the first and second recording areas 61 and 62 of the recording means 55.
When 3: 2 pulldown information exists in both
Of the first and second recording areas 61 and 62.
3: 2 pulldown as it is during the period when the information is shared
Information is stored in the recording means 63, and the first and second recording
The 3: 2 pulldown information of the areas 61 and 62 is not common.
3: 2 pulldown of 3: 2 pulldown information
To the recording means 63
Let me save. First and second recording areas 6 of recording medium 56
The bit stream files recorded in 4, 65 are
Decoded and expanded by a decoding circuit (not shown)
3: 2 pulldown detection again as single angle original image signal a4
The circuit 52 outputs the 3: 2 pull-down information d4 of the recording means 63.
Using 3: 2 pulldown and field-frame conversion
The compression is performed, and the image signal c4 is sent to the compression encoding circuit 54.
Supplied and recorded in the first recording area 64 of the recording medium 56.
It is. With such a configuration, 3: 2 pull-down
The information extraction circuit 57 performs compression encoding by the compression encoding circuit 54.
Time code information and 3:
Extract and save 2 pulldown phase information 3: 2 pull
Down information extraction and storage means, and recording means 55 and 63
And 3: 2 pull-down information comparison and generation circuit 58
Time code information from 3: 2 pull-down information extraction means
Are the same, there is a plurality of 3: 2 pulldown phase information.
In some cases, multiple 3: 2 pulldown phase information is common
In the period where
3: 2 pull-down information created and saved in
3: 2 pull-down during period when 3: 2 pull-down phase information is different
The pull-down phase of 3: 2 of the down phase information is changed to the pull-down phase of 2
3: 2 pull-down information creation and storage means
3: 2 pull-down information comparison storage means
3: 2 pull-down detection circuit 52
3: 2 pulldown information from down information comparison storage means
Therefore, the 3: 2 format of the original image signal of the NTSC system is once again used.
3: 2 program that performs
Down means. In such an embodiment of the invention, first,
The first original image signal a4 for multi-angle is input terminal 5
When input from 1, 3: 2 pull-down detection circuit 5
2 is 3: 2 pulldown detection for original image signal a4
Then, 3: 2 pull-down is performed based on the detection result,
Create an image signal c4 and record it via the compression encoding circuit 54
A video bit stream file is generated on the medium 56.
Next, the original image signal a that has been edited for multi-angle
4 is input from the input terminal 51, the 3: 2 pull
The circuit detection circuit 52 outputs a 3: 2 signal to the original image signal a4.
Down, and a 3: 2 pull-down
The compression encoding circuit 54
Video bit in the second recording area 65 of the recording medium 56 via the
Generate stream file. First and second recording areas 6 of recording medium 56
Unedited multi-angle video video recorded at 4,65
Bitstream with edited streams and edited video
Are input to the 3: 2 pull-down information extraction circuit 57, respectively.
Is done. The 3: 2 pulldown information extraction circuit 57
Code and 3: 2 phase information for each video bit stream.
Pull out of the ream,RecordFirst and second means 55
The recording areas 61 and 62 are generated. As a result, 3: 2
The down-down information comparison and generation circuit 58 performs first and second recording.
Period when 3: 2 pulldown information of areas 61 and 62 is common
Between the 3: 2 pulldown information as it is to the recording means 63
Is saved, and 3: 2 of the first and second recording areas 61 and 62 are stored.
3: 2 pull during periods when pull-down information is not common
Pull down 3: 2 pull down phase of down information to 2
The phase is changed and stored in the recording means 63. Next, from the input terminal 51, a multi-angle
Again input the unedited original image signal a4 of
3: 2 pulldown detection mode selection signal b4 is false,
The 3: 2 pull-down detection circuit 52 detects the original image signal a4.
And use 3: 2 pull-down information d4 from recording means 63
To determine the 3: 2 phase and perform 3: 2 pulldown
A signal c4 is created and recorded on the recording medium via the compression / encoding circuit 54.
The video bit stream file is stored in the first recording area 64 of the body 56.
Generate a file. Below, the rest for multi-angle
A similar process is performed on the original image signal a4 that has not been collected.
The original image signal read and decoded from the recording means 56
By performing the same processing for the signal a4, the recording
3: 2 using 3: 2 pulldown information d4 from stage 63
3: 2 pulldown operation and compression encoding operation with phase determined
Operation for all multi-angle original image signals.
Next. By the above operation, the embodiment of the invention shown in FIG.
The same effect as in the embodiment is obtained. [0083] According to the present invention, an original for a multi-angle is provided.
When compressing and encoding a 3: 2 pulldown image signal
Can prevent the number of pictures in the GOP from differing,
The amount of redundant code can be reduced, and the image quality can be improved.
You.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of a moving image compression encoding apparatus according to the present invention. FIG. 2 is an explanatory diagram illustrating compression encoding of an original image signal for multi-angle in FIG. 1; FIG. 3 is an explanatory diagram showing how the moving picture compression encoding apparatus of FIG. 1 converts a 24 Hz progressive scanning signal, which is a movie material, into an NTSC signal. FIG. 4 is an explanatory diagram illustrating an outline of an example of a 3: 2 pull-down detection method by the moving image compression encoding device of FIG. 1; FIG. 5 is a block diagram showing a second embodiment of the moving picture compression encoding apparatus according to the present invention. FIG. 6 is a flowchart showing an algorithm of the operation of the 3: 2 pull-down information extraction circuit in FIG. 5; FIG. 7 is a block diagram showing a third embodiment of the moving picture compression encoding apparatus according to the present invention. FIG. 8 is an explanatory diagram showing a pull-down phase at which the 3: 2 pull-down information comparison and generation circuit in FIG. 7 makes a determination. FIG. 9 is a flowchart showing a specific algorithm example of the 3: 2 pull-down information comparison and generation circuit in FIG. 7; FIG. 10 is a block diagram showing a fourth embodiment of the moving picture compression encoding apparatus according to the present invention. [Description of Signs] 12 3: 2 Pulldown Detection Circuit 14 Compression Encoding Circuit 15 Recording Means 16 Recording Medium

──────────────────────────────────────────────────続 き Continued on front page (58) Field surveyed (Int.Cl. ⁷ , DB name) H04N 5/76-5/956 H04N 7/24

Claims (1)

(57) [Claim 1] Each frame of a progressively scanned video signal having a frame frequency of 24 Hz is obtained by repeating two fields of the same image.
Return image or repeated image of 3 fields of the same image
A moving image compression encoding apparatus that receives an NTSC original image signal generated by conversion as an input, detects a repetitive image of the input NTSC original image signal, and detects a repetitive image. A 3: 2 pull-down detection circuit that creates and outputs an image signal by performing 3: 2 pull-down and further performs field-frame conversion, and a compression that compresses and encodes the image signal output by the 3: 2 pull-down detection circuit An encoding circuit; a 3: 2 pull-down information generating and storing means for generating and storing pull-down information which is a detection result of the 3: 2 pull-down detection circuit; and a 3: 2 detection result of the 3: 2 pull-down detection circuit.
When there are a plurality of pieces of pull-down information, the 3: 2 pull-down information is stored in the 3: 2 pull-down information creating and storing means as it is during a period in which the plurality of 3: 2 pull-down information is common, and the plurality of 3: 2 pull-down information is different. In the period, the 3: 2 pull-down information is changed from the 3: 2 pull-down phase of the 3: 2 pull-down information to the pull-down phase of 2, and is stored in the 3: 2 pull-down information creating and storing means. A moving picture characterized by comprising: a 3: 2 pull-down means for performing a 3: 2 pull-down of the NTSC original image signal again according to the 3: 2 pull-down information from the comparison storage means and supplying the same to the compression encoding circuit. Image compression encoding device.