JPH11313331A - Image signal processing system, decoder, image signal processing method and decoding method - Google Patents

Abstract

(57) [Summary] [PROBLEMS] To sufficiently suppress image quality degradation in re-encoding using compression encoding. SOLUTION: An MPEG decoder 10 is supplied with an encoded bit stream as an image signal, decodes the bit stream to generate a decoded image signal, and supplies the decoded image signal to a multiplexer 11. The coding parameters including the coding bit amount and the average quantization scale used at the time of the first coding) are transmitted to the multiplexer 11 and the terminal 16a.
To supply. The multiplexer 11 superimposes these on the decoded image signal. The output of the multiplexer 11 is supplied to a separator 13 via a recording / reproducing system, where the encoding parameters are separated and supplied to a terminal 16b. On the other hand, the image signal is
To the MPEG encoder 14. Switch 1
By turning 5 to the terminal 16b side, re-encoding can be performed under the encoding parameters used in the first encoding.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image signal processing system, a decoder, an image signal processing method, and a decoding method that are particularly suitable for applications requiring high image quality.

[0002]

2. Description of the Related Art In an editing process for an image signal encoded by the MPEG system, when an encoded image signal is decoded, and when the decoded image is re-encoded, a normal video signal is encoded. In some cases, the image quality is significantly deteriorated. As a cause of such image quality deterioration, various encoding parameters such as a picture type and a motion vector do not match before and after re-encoding, that is, when encoding is first performed by the MPEG method and when re-encoding is performed. is there.

The case where the picture type is not matched among the coding parameter mismatches will be described with reference to FIG. FIG. 9A shows a GOP when n = 15.
(GroupOf Pictures) An example of a picture type for one input decoded image, that is, a decoded image to be re-encoded is shown. When the picture type phase lock is performed as shown in FIG. 9B, the reference picture at the time of re-encoding is an I picture in the input decoded picture shown in FIG. 9A (that is, at the time of first encoding). The thing is used as it is.

On the other hand, when the phase lock is not performed as shown in FIG. 9C, a B picture having a high degree of image quality deterioration, such as a third picture, is used as a reference picture at the time of re-encoding. Will be used. As a result, the accuracy of re-encoding is reduced, and large image quality degradation occurs.

[0005] Also, in the case of using a value calculated by performing recalculation at the time of re-encoding without using the same value before and after re-encoding for other encoding parameters, for example, a motion vector, the prediction accuracy is reduced. The accuracy of re-encoding decreases.

On the other hand, if all the encoding parameters including the picture type and the motion vector are matched before and after the re-encoding, the image quality hardly deteriorates due to the re-encoding. However, since the amount of information to be handled increases as the number of encoding parameters increases, it is not practical to match all encoding parameters before and after re-encoding. In general, it is known that if the picture type or the motion vector is matched before and after re-encoding, it is possible to reduce image quality degradation by a considerable amount.

[0007]

However, when high image quality is required, even if the picture type and the motion vector are matched before and after re-encoding, the suppression of image quality deterioration is not sufficient.

[0008] A motion vector expresses motion information in macroblock units, and the amount of information is relatively large. For this reason, from the viewpoint of improving the processing efficiency at the time of re-encoding, by supplying the motion vector to the encoder that performs the re-encoding, the image quality can be degraded even when a configuration that matches the motion vector before and after the re-encoding cannot be adopted. A coding parameter that is effective for suppression and has a smaller amount of information than a motion vector is required.

Therefore, an object of the present invention is to provide an image signal processing system, a decoder, an image signal processing method, and a decoding method, which can sufficiently suppress image quality deterioration and require a small amount of information to be transmitted. It is in.

[0010]

An image signal processing system according to the present invention generates an encoded bit amount and / or an average quantization scale as representative values of encoding parameters of an input encoded image signal. A decoder for decoding the input coded image signal to generate a decoded image signal, and outputting the coding parameter together with the generated decoded image signal; and And an encoder for performing image encoding.

Further, the image signal processing system according to the present invention decodes an input coded image signal to generate a decoded image signal, and generates a decoded image signal in a signal format for transmitting the decoded image signal. In the signal part corresponding to the invalid period,
A decoder that superimposes the coding parameters of the input coded image signal and outputs the coding parameter together with the decoded image signal, and uses the coding parameter to perform image coding on the decoded image signal. It is characterized by having an encoder to perform.

[0012] A decoder according to the present invention comprises: a coding parameter generating means for generating a coding bit amount and / or an average quantization scale as a representative value of coding parameters of an input coded image signal; Decoding means for decoding the encoded image signal to generate a decoded image signal, and outputting the encoding parameter generated by the encoding parameter generating means together with the decoded image signal generated by the decoding means. Features.

Further, a decoder according to the present invention decodes an input coded image signal to generate a decoded image signal, and a signal format for transmitting the decoded image signal generated by the decoding means. Output means for superimposing the coding parameter of the input coded image signal on the signal portion corresponding to the invalid period in the output, and outputting the coding parameter together with the decoded image signal. .

In the image signal processing method according to the present invention, a coding bit amount and / or an average quantization scale is generated as a representative value of coding parameters of an input coded image signal. Decoding the image signal to generate a decoded image signal, outputting the encoding parameter together with the generated decoded image signal, and performing image encoding on the decoded image signal using the encoding parameter. And

Further, the image signal processing method according to the present invention comprises:
The input coded image signal is generated by decoding the input coded image signal to generate a decoded image signal, and a signal portion corresponding to an invalid period in a signal format for transmitting the generated decoded image signal. And outputting the coding parameter together with the decoded image signal, and performing image coding on the decoded image signal using the coding parameter.

In the decoding method according to the present invention, a coding bit amount and / or an average quantization scale is generated as a representative value of a coding parameter of an input coded image signal. Is decoded to generate a decoded image signal, and the encoding parameter is output together with the decoded image signal.

Further, the decoding method according to the present invention decodes an input coded image signal to generate a decoded image signal, and corresponds to an invalid period in a signal format for transmitting the generated decoded image signal. The coding parameter of the input coded image signal is superimposed on the signal portion to be output, and the coding parameter is output together with the decoded image signal.

According to the invention, the encoding parameter used at the time of the first image encoding is supplied to the encoder that performs the re-encoding. Therefore, re-encoding can be performed using the same encoding parameters as those used in the first inter-image compression encoding.

In particular, by supplying an encoding bit amount at the time of the first encoding, an average quantization scale, and the like as encoding parameters to an encoder that performs re-encoding,
It is possible to perform control to match the coded bit amount before and after re-encoding.

[0020]

Embodiments of the present invention will be described below in detail with reference to the drawings.

First, the basic concept of the present invention will be described.

As a result of a study including an experiment and the like on the effects of various coding parameters on image quality deterioration, it was found that a difference in the amount of coded bits before and after re-encoding is one of the causes of image quality deterioration. The reason that the encoded bit amount differs before and after re-encoding is that the rate control at the time of the first encoding and the rate control at the time of re-encoding are different.

For example, in a feed-forward type rate control, the rate control is performed according to the degree of difficulty in coding an image. In this case, in an image that has been once decoded by the MPEG method and decoded, the information amount of the image that was a B picture at the time of the first encoding is smaller than that of an I picture or a P picture. It is determined that the encoding difficulty of the picture (a parameter indicating the ease of compression as described later) is low. For this reason, when re-encoding by matching the picture type before and after re-encoding, image quality is deteriorated as a result of performing control to reduce the bit amount of the B picture.

Therefore, as will be described later, by controlling the encoded bit amounts before and after re-encoding so as to match, it is possible to suppress image quality deterioration. Furthermore, as a result of further study, it was found that matching the amount of encoded bits before and after re-encoding is effective even when the picture types are matched and the motion vectors are not matched.

In order to match the coded bit amount before and after re-encoding, there are a method of supplying the coded bit amount itself to an encoder that performs re-encoding, and a method of supplying a quantization scale to such an encoder. Of course, both of them may be supplied to such an encoder. Since the quantization scale itself has a relatively large amount of information, in the present invention, an average quantization scale is used in place of the quantization scale in consideration of processing efficiency and simplicity.

According to the above description, combinations of encoding parameters that can be used for suppressing image quality degradation during re-encoding are as follows. These encoding parameters are used in encoding / decoding in units of pictures or slices.

Picture type, motion vector, coding bit amount, average quantization scale Picture type, motion vector, coding bit amount Picture type, motion vector, average quantization scale Picture type, coding bit amount, average quantization scale Picture type, coded bit amount Picture type, average quantization scale Here, the coded bit amount is the sum of the code or bit amount for each picture or slice. This value may be a value represented by the number of bytes. Also, for example, 128 bytes (1 kbi
The value may be rounded in units of t).

The average quantization scale is an average value of quantization scales of MBs (macroblocks) having significant DCT coefficients in units of pictures or slices. MB with DCT coefficient is
This is an MB excluding the skipped MB and not coded MB defined by MPEG.

When the coded image signals are all I pictures, there is no motion vector, and there is no need to input the picture type to the encoder.

With reference to FIG. 1, an embodiment of the present invention for realizing, for example, one of the above-described combinations will be described. The MPEG decoder 10 is supplied with a bit stream as an image signal encoded by the MPEG method.
Then, the bit stream is subjected to a decoding process prior to re-encoding to generate a decoded image signal.
Feed to 1. The MPEG decoder 10 also calculates the picture type, the motion vector, the coded bit amount of the picture, and the average quantization scale used at the time of decoding (therefore, at the time of the first encoding for generating the above-described bit stream). The signal is supplied to the multiplexer 11 and the terminal 16a.

The multiplexer 11 superimposes these encoding parameters on the decoded image signal. Specifically, these encoding parameters are embedded in a predetermined transmission signal format, for example, header information of a decoded image signal defined by a 525 line / 60 field system, an empty area of the decoded image signal such as blanking, or the like. can do. When embedding in blanking, for example, VB-I
User defined areas in D format can be used.

The output of the multiplexer 11 is supplied to a recording / reproducing system including a recording medium 12 such as a magnetic tape.
Will be recorded. On the reproduction side, a signal reproduced from the recording medium 12 is supplied to the separator 13. The separator 13 separates the encoding parameter embedded in the supplied signal and supplies it to the terminal 16b, and also outputs the image signal to the MPE.
This is supplied to the G encoder 14. Then, when the switch 15 is turned to the terminal 16b side, the encoding parameters from the separator 13 are supplied to the MPEG encoder 14 that performs re-encoding. This allows re-encoding to be performed using the encoding parameters used during decoding (and thus during the first encoding to generate the bit stream described above).

On the other hand, according to the configuration of FIG. 1, the encoding parameters used at the time of decoding (therefore, at the time of the first encoding for generating the above-mentioned bit stream) are directly transmitted to the MPE.
It is also possible to supply to the G encoder 14 and perform re-encoding using such encoding parameters. In this case, control may be performed so that the switch 15 is moved to the terminal 16a side. In FIG. 1, each of the encoding parameters is output from the MPEG decoder 10. However, the configuration in which each of the encoding parameters is extracted from the bit stream and each of the extracted encoding parameters is output is described in the MPEG decoder 20. May be provided separately.

Next, another embodiment of the present invention for realizing the above-mentioned combination will be described with reference to FIG.
A bit stream as an image signal encoded by the MPEG system is supplied to an MPEG decoder 20, a motion vector and picture type extracting circuit 22, and an encoded bit amount counter 23. The MPEG decoder 20 decodes the supplied bit stream to generate a decoded image signal. Then, the decoded image signal is supplied to the MPEG encoder 21.

The motion vector and picture type extracting circuit 22 extracts a motion vector and a picture type from a bit stream, and supplies the extracted motion vector and picture type to the MPEG encoder 21. Similarly, the coded bit amount counter 23 calculates the coded bit amount from the bit stream, and supplies the calculated coded bit amount to the MPEG encoder 21.
The MPEG encoder 21 re-encodes the decoded image signal supplied from the MPEG decoder 20 using the motion vector, the picture type, and the coded bit amount supplied in this manner.

Next, another embodiment of the present invention for realizing the above-mentioned combination will be described with reference to FIG. A bit stream as an image signal encoded by the MPEG system is converted into an MPEG decoder 30, an average quantization scale extracting circuit 32, a motion vector and picture type extracting circuit 33, and an encoded bit amount counter.
Supplied to The MPEG decoder 30 decodes the supplied bit stream to generate a decoded image signal. Then, the decoded image signal is supplied to the MPEG encoder 31.

The average quantization scale extracting circuit 3
2 extracts the quantization scale from the bit stream, calculates the average quantization scale for each picture,
The value is supplied to the MPEG encoder 31. On the other hand, the motion vector and picture type extracting circuit 33 extracts a motion vector and a picture type from the bit stream, and supplies the extracted motion vector and picture type to the MPEG encoder 31.

Similarly, the coded bit amount counter 34
Calculates the amount of coded bits from the bit stream and MPE
This is supplied to the G encoder 31. MPEG encoder 31
Re-encodes the decoded image signal supplied from the MPEG decoder 30 using the average quantization scale, the motion vector, the picture type, and the coded bit amount supplied in this manner.

In FIG. 2 and FIG. 3, a circuit for extracting each encoding parameter from the bit stream is an MPEG circuit.
It is separate from the decoder 20. In contrast,
Since each encoding parameter is also obtained in the decoding process for the bit stream, the MPEG decoder 20 may be configured to output each encoding parameter.
If the bit amount of any of the coding parameters becomes a burden in transmission, the coding parameters such as the coding bit amount and the average quantization scale are quantized to such an extent that image quality does not matter. May be transmitted. This makes it possible to reduce the total information amount of the coding parameters to be transmitted.

The above-described embodiment of the present invention realizes or among the above-described combinations. However, the MPEG decoder 10 outputs the encoding parameters corresponding to the above-mentioned combinations of the above. And MP
If the EG encoder 14 is configured to perform encoding using these encoding parameters, it is possible to realize a combination of the following. In particular, in the configuration that realizes the combination of the above, the transmission of the motion vector is not required, so that the total information amount of the transmitted coding parameters can be reduced.

The re-encoding procedure in the above-described embodiment of the present invention will be described with reference to the flowchart of FIG. Here, the description will be made on the assumption that the combination of the encoding parameters is used. However, when another combination is used, only the type of the encoding parameter to be handled is changed, and the re-encoding can be performed in substantially the same procedure.

In step S1, the amount of coded bits is calculated from the bit stream. In step S2, a motion vector and a picture type are extracted from the bit stream to encode bit amounts. Then, in step S3, the assigned bit amount is set as the coding bit amount, and encoding is performed using the extracted motion vector and picture type. Further, in step S4, it is determined whether or not the encoding in step S3 has been completed for all the frames. If it is determined in step S4 that encoding has been completed for all frames, re-encoding is terminated. On the other hand, if it is determined in step S4 that the encoding has not been completed for all the frames, the process proceeds to step S1 to perform processing on the remaining frames.

FIG. 5 shows a procedure of feedforward rate control performed using the encoding parameters supplied to the encoder that performs re-encoding.
This will be described with reference to the flowchart of FIG. Here, the description will be given on the assumption that the combination of the encoding parameters is used. However, even when another combination is used, re-encoding can be performed in substantially the same procedure only by changing the type of the encoding parameter to be handled.

After the encoding process starts, first, in step S
In 101, a variable pict_i representing the picture order is set to 1
And Thus, the processing for the first picture is started. In step S102, the bit amount allocated to the pict_i-th picture is set to the original coded bit amount. Next, in step S103, the variable mb_i representing the order of the macroblocks in the picture is set to 1, and the processing for the first macroblock is started. Next, in step S104, an initial value of the learning encoding parameter is calculated from the average quantization scale according to the following equation (1). The learning coding parameters are used in the calculation of the quantization scale (Step S106).

[0045]

(Equation 1)

Here, KQ [1] is the initial value of the learning coding parameter, and MQT is the average quantization scale. XA and KR are respectively expressed by the following equation (2):
It is calculated according to (4).

[0047]

(Equation 2)

Here, target_bit is the allocated bit amount per picture, which is set to the original coded bit amount in step S102. The encoding difficulty x_mb [mb_i] for the mb_i-th picture is calculated according to the following equation (3).

[0049]

(Equation 3)

Here, fbit [mb_i] is a predicted value of the amount of generated bits when the MB is coded on a fixed quantization scale. The calculation of fbit [mb_i] is I
For a picture, this is performed based on the average separation residual of the image, and for P and B pictures, it is performed based on the prediction residual between the images. Also, weight_mb [mb_
i] is a parameter indicating the degree of conspicuousness of image deterioration, and is determined from, for example, the flatness, brightness, redness, and the like of the image. NMB is the total number of macroblocks in the frame.

[0051]

(Equation 4)

Here, frame_rate is the number of frames per second.

Further, in step S105, mb_i
The allocated bit amount for the macro block is calculated according to the following equation (5).

[0054]

(Equation 5)

Here, target_bit_mb [m
b_i] is the mb_i-th assigned bit amount. Further, the encoding difficulty x_mb [mb_i] is calculated according to the above equation (3).

Next, in step S106, the quantization scale for the mb_i-th macroblock is calculated according to the following equation (6).

[0057]

(Equation 6)

Here, mQ [mb_i] is the quantization scale for the mb_i-th macroblock, and KR, XA, and weight_mb [mb_i] are as described above. Further, the mb_i-th learning coding parameter KQ [mb_i] is calculated in step S10 described above.
4 and is updated by the later-described step S108 along with the progress of the processing.

In step S107, the mb_i-th macroblock is quantized and encoded. And
The process proceeds to step S108, and the learning encoding parameter is updated according to the following equation (7).

[0060]

(Equation 7)

Here, frame_bit_gener
added [mb_i] is the total amount of coded bits up to the mb_i-th macroblock of the current picture,
um_target_bit_mb [mb_i] is m
This is the total allocated bit amount up to the b_i-th macroblock.

Next, in step S109, the number of the macroblock is updated by adding 1 to mb_i. Then, in step S110, mb_i
Is larger than the total number NMB of macroblocks per picture. If it is determined that mb_i is larger than NMB, it is determined that the processing for the picture has been completed, and the flow shifts to step S111. On the other hand, m
When it is determined that b_i is equal to or smaller than NMB, the process proceeds to step S105, and the subsequent (that is, mb_i is 1).
Start processing for the (increased) macroblock.

In step S111, 1 is added to pict_i indicating the above-mentioned picture number. Further, the processing shifts to step S112, and it is determined whether or not the number of pictures using pict_i as a unit of processing, for example, the total number of pictures NPICT per GOP is larger. If it is determined in step S102 that pict_i is equal to or smaller than NPICT, the process shifts to step S102 to start processing for a subsequent picture (that is, pict_i has increased by 1). On the other hand, if it is determined in step S112 that pict_i is larger than NPICT, it is determined that the processing for the processing unit such as the GOP has been completed, and the encoding ends.

In the above-described rate control procedure, the mb_i-th quantization scale related to the processing accuracy of the mb_i-th macroblock is calculated in step S106. Then, for this calculation, the value of the learning coding parameter is used (see equation (6)). The learning encoding parameter is updated in step S108 (see equation (7)), but the initial value is calculated from the average quantization scale in step S104 (see equation (1)).

The value of the average quantization scale used at this time is the value used at the time of the first encoding transmitted by the above-described configuration. By such processing, appropriate rate control is performed based on the value of the average quantization scale used at the time of the first encoding, and as a result, control to match the encoded bit amount before and after re-encoding is performed. It is possible.

In the above description, the case where the coded bit amount of the picture is matched before and after the re-encoding has been described. Next, the rate control method for changing the coded bit amount by performing the re-encoding. Will be described. The flow of the process is the same as that of the flowchart of FIG. 5, and the details of the process in step S102 and step S104 are different.

That is, for example, assume that the bit rate of the original coded bit stream is R1, and the bit rate of the coded bit stream created by re-encoding it is R1.
When it is set to 2, step S10 in the flowchart of FIG.
The setting of the number of bits to be allocated to the picture in 2 is performed based on the following equation (8).

[0068]

(Equation 8)

Here, picture_bit_siz
e is the coding bit amount of the picture of the original coding pit stream, and target_bit is the assigned bit amount of the picture to be re-encoded.

Further, step S in the flowchart of FIG.
The initial value of the learning coding parameter at 104 is set as follows. This replaces the calculation of KQ [1] in equation (1) above.

[0071]

(Equation 9)

Where MQT is the average quantization scale of the original coded pit stream. Other variables have the same meanings as the variables described above.

Note that in the case of MPEG, there are three types of macroblocks, similar to the picture type. That is, an intra-coded (Intra) macroblock, a forward (Forward) inter-frame prediction macroblock that predicts the future from the past, and a backward (Bac) macroblock that predicts the past from the future.
kward) There are an inter-frame prediction macro block and an interpolative macro block predicting from both forward and backward directions.

All macroblocks in an I picture are
This is an intra-coded macroblock. The P picture includes an intra-frame coded macro block and a forward inter-frame prediction macro block. The B picture includes all four types of macroblocks described above.

More specifically, the present invention relates to a method for converting an input analog video signal into a digital video signal and compressing the converted digital video signal, or compressing a directly input digital video signal and then compressing the digital video signal. The present invention can be applied to a video signal recording / reproducing apparatus which records and expands and reproduces a compressed digital video signal recorded on the optical disk.

First, the configuration and operation of the recording processing system will be described. In FIG. 6, a digital video signal is directly supplied to an input terminal 61, and an analog video signal is supplied to an input terminal 62. The analog video signal is an image signal, a broadcast video signal received by an antenna, or the like. The analog video signal is converted into a digital video signal by the A / D converter 63. The digital video signal from the input terminal 61 and the digital video signal from the A / D converter 63 are supplied to the MPEG encoder 65 via the video signal controller 64.

The video signal control unit 64 controls the A / A in accordance with the control of the system controller 75 based on the information obtained via the recording control signal input unit 74 in accordance with the setting of the user.
Digital video input from D conversion unit 63, input terminal 61
And selects one of the digital video input signals supplied from the above and supplies it to the MPEG encoder 65. The MPEG encoder 65 performs MPEG compression coding on the video signal from the video signal control unit 64.

The address of the digital video signal compressed by the MPEG encoder 65 is specified by a memory control section 67 controlled by a system controller 75 via a bus, and the digital video signal is stored in a recording buffer memory section 66 a of an integrated buffer memory 66. Is stored.

The digital video signal stored in the recording buffer memory 66a is supplied to the optical disk drive via the bus, the data processor 68 and the recording / reproduction switch 69. The data processing unit 68 includes a recording signal processing unit 68a and a reproduction signal processing unit 68b. The recording signal processing section 68a performs processing such as error correction encoding and digital modulation, and the reproduction signal processing section 68b performs processing such as error correction and demodulation of digital modulation. The optical disk drive irradiates an optical disk 71 with a recording laser beam to record a signal, and irradiates a reproduction laser beam to reproduce a signal, and a spindle motor for rotating and driving the optical disk 71. 72. The optical head 70 and the spindle motor 72 are controlled by a disk / head controller 73. The optical head 70 converts the output signal of the recording signal processor 68a into an optical disk 71.
Will be recorded. The optical disk 71 is rewritable, and an MO (magneto-optical) disk, a phase change type disk, or the like can be used.

The system controller 75 controls the optical disk drive via the disk / head control unit 73 and also manages the state of the optical disk drive. The system controller 75 transmits the information to the memory control unit 67 and transmits the information to the integrated buffer memory 66. The control of the data supply is performed.

Next, the reproduction processing system will be described. The reproduction processing system includes an MPEG decoder 77 for decoding a reproduction signal supplied from the reproduction system buffer memory unit 66b of the integrated buffer memory 66 via the bus, and an MPEG decoder 7
Video signal controller 64 for switching the decoded video signal from
And a D / A converter 79 for converting the video signal switched by the video signal controller 64 into an analog video signal.

In the reproduction mode, the optical disk drive is controlled by the disk / head control unit 73 to control the servo and head movement, and reproduces the reproduction signal via the reproduction signal processing unit 68b of the data processing unit 68 and the reproduction system buffer memory via the bus. Part 6
6b. The reproduction-system buffer memory unit 66b includes:
The reproduction video signal is supplied to the MPEG decoder 77 while balancing the writing and reading of the reproduction signal.
The MPEG decoder 77 performs an MPEG decoding process on the reproduced signal, and supplies the decoded video signal to the video signal control unit 64.

The video signal control section 64 is controlled by a system controller 75 based on information obtained via a reproduction control signal input section 76 in accordance with a user's setting, and is controlled by the MPEG.
Switching processing is performed on the decoded video signal from the decoder 77, and the decoded video signal is output to the D / A converter 79 or the output terminal 78. The D / A converter 79 converts the digital video signal switched and controlled by the video signal controller 64 into an analog video signal, and converts the analog video signal into an output terminal 78.
Output to

The video signal control section 64 comprises a changeover switch SW1 and a changeover switch SW2. The changeover switch SW1 has an input terminal a to which a digital video signal input from the input terminal 61 is supplied, and an A / D converter 63
An input terminal b to which a digital video signal from
An input terminal c to which a decoded video signal from the MPEG decoder 77 is supplied and an output terminal d to input a video signal to the MPEG encoder 65 are provided. The changeover switch SW2 has an input terminal e to which a digital video signal from the input terminal 61 is supplied, an input terminal f to which a decoded video signal from the MPEG decoder 77 is supplied, and an output terminal
8 and an output terminal g for outputting a video signal to the D / A converter 79.

The switching of the switches SW1 and SW2 in the video signal control section 64 is controlled by the system controller 75. More specifically, if the command from the user supplied to the system controller 75 via the recording control signal input unit 74 specifies the external digital video signal from the input terminal 61, the switch S
The input terminal a of W1 is connected to the output terminal d. If the video signal from the input terminal 62 is designated, the input terminal b is connected to the output terminal d.

Further, if there is a command to connect any video input and video data reproduced by the optical disk drive to edit and record it again on the optical disk 71, the system controller 75 connects the output terminal d to the input terminal c. Control the timing. That is, the MPEG decoder 77
Is supplied directly to the MPEG encoder 65 for re-encoding.

The video signal recording / reproducing apparatus includes an integrated buffer memory 66 for changing the allocation of storage areas for the recording system and the reproduction system, and a storage area allocation for the integrated buffer memory 66 according to the recording mode or the reproduction mode. The processing is controlled by the system controller 75. That is, the areas of the recording-system buffer memory unit 66a and the reproduction-system buffer memory unit 66b are variable under the control of the system controller 75 via the memory control unit 67. For example, at the time of recording, the recording system buffer memory 66a
Occupies all of the integrated buffer memory 66. At the time of reproduction, the reproduction system buffer memory unit 66b occupies all. In addition, at the time of simultaneous recording and reproduction, half of the memory capacity may be secured.

Further, the re-encoding according to the present invention may be applied only near the editing point, and in other parts, an editing process for switching the bit stream may be performed.

Next, the effect of the re-encoding according to the present invention will be described with reference to the conventional re-encoding. The comparison is based on the standard five sequences (bus, bicycle, mobil & calender, flower) for examining the degree of image quality deterioration according to the standards of the International Council for Wireless Communications (CCIR).
This is performed using SNR (Signal Nioze Rate) when re-encoding is performed for garden and cheerleader. For an image signal of 8 bits per pixel, the SNR is calculated according to the following equation (10).

[0090]

(Equation 10)

Here, pixel_num is the number of pixels per picture. Also, MeanError is 1
This is the average number of errors per pixel, and is expressed by the following equation (11).
Is calculated according to

[0092]

[Equation 11]

Here, SumError is the total number of errors for all pixels, and is calculated according to the following equation (12).

[0094]

(Equation 12)

[0095] Here, Org _i is the i th pixel value of the original image. Cur _i is the i-th pixel value of the encoded image.

FIG. 7 shows the conventional re-encoding in which the picture type and the motion vector are matched between the re-encoding and the first encoding, and the re-encoding of the coded bit amount in addition to the picture type and the motion vector. An example of the comparison of the SNR is shown for the re-encoding according to the present invention and the re-encoding according to the present invention, which is made to coincide with the time of the first encoding.
For any of the five sequences, the SNR when re-encoding according to the present invention is higher shows a higher value, which indicates that the present invention is effective in suppressing image quality deterioration.

FIG. 8 shows the conventional re-encoding in which only the picture type is matched at the time of re-encoding and the first encoding, and the encoding bit amount in addition to the picture type and the initial encoding. Regarding re-encoding according to the present invention, which is matched at the time of encoding,
An example of NR comparison was shown. In FIG. 8 as well, the SNR when re-encoding according to the present invention is performed for any of the five sequences shows a higher value, indicating that the present invention is effective in suppressing image quality degradation.

[0098]

As described above, according to the present invention, at the time of re-encoding, a coding bit amount and / or an average quantization scale are generated as representative values of coding parameters of an input coded image signal. Decoding the input encoded image signal to generate a decoded image signal, outputting the encoding parameter together with the generated decoded image signal, and using the encoding parameter to perform image encoding on the decoded image signal. Is performed, the coding parameters such as the coding bit amount and the average quantization scale used at the time of re-encoding can be made to match those used at the time of the first encoding. As a result, the amount of encoded bits can be matched between re-encoding and initial encoding.

In general, the rate control in encoding differs depending on the degree of difficulty in encoding an image, and as a result, the amount of encoded bits fluctuates. Such a change in the amount of coded bits causes image quality degradation in re-encoding. Therefore, image quality deterioration can be suppressed by the above-described processing for matching the coded bit amounts at the time of re-encoding and the first encoding.

In addition, matching the amount of encoded bits in this way is effective in suppressing image quality degradation even when the picture types are matched and the motion vectors are not matched. In consideration of this point, depending on the required image quality, transmission of a motion vector for matching a motion vector before and after re-encoding may not be performed. In such a case, the total bit amount for transmitting the encoding parameters can be reduced, and the configuration related to the transmission can be simplified and the cost can be reduced. This can contribute to cost reduction.

[Brief description of the drawings]

FIG. 1 is a block diagram for describing an embodiment of the present invention.

FIG. 2 is a block diagram for explaining another embodiment of the present invention.

FIG. 3 is a block diagram for explaining still another embodiment of the present invention.

FIG. 4 is a flowchart for explaining a re-encoding procedure according to the present invention.

FIG. 5 is a flowchart for explaining a rate control procedure in re-encoding according to the present invention.

FIG. 6 is a block diagram for describing an image signal processing system to which the present invention can be applied.

FIG. 7 is a schematic diagram for explaining an example of the effect of the present invention.

FIG. 8 is a schematic diagram for explaining another example of the effect of the present invention.

FIG. 9 is a schematic diagram for explaining a problem when picture types do not match before and after re-encoding.

[Explanation of symbols]

1,11 multiplexer, 13 separator, 14, 21, 31
MPEG encoder, 23, 34 Encoding bit amount counter, 32 Average quantization scale extraction circuit

Claims (24)

[Claims] An encoded bit amount and / or an average quantization scale is generated as a representative value of an encoding parameter of an input encoded image signal, and the input encoded image signal is decoded to decode a decoded image. A decoder that generates a signal and outputs the encoding parameter together with the generated decoded image signal; and an encoder that performs image encoding on the decoded image signal using the encoding parameter. Image signal processing system. 2. The decoder outputs a picture type and / or a motion vector together with the decoded image signal as the encoding parameter, in addition to an encoded bit amount and / or an average quantization scale. The image signal processing system according to claim 1. 3. The image signal processing system according to claim 1, wherein the decoder outputs, on a picture basis, an encoding parameter used when performing image encoding together with the decoded image signal. 4. The image signal processing system according to claim 1, wherein the decoder outputs, in units of slices, an encoding parameter used when performing image encoding together with the decoded image signal. 5. The image signal processing system according to claim 1, wherein said encoder performs rate control based on an input coded bit amount and / or an average quantization scale. 6. A decoded image signal is generated by decoding an input coded image signal, and a signal portion corresponding to an invalid period in a signal format for transmitting the generated decoded image signal is provided. A decoder that superimposes the coding parameters of the input coded image signal and outputs the coding parameter together with the decoded image signal; and performs image coding on the decoded image signal using the coding parameter. An image signal processing system comprising an encoder. 7. The decoder generates a coding bit amount and / or an average quantization scale as a representative value of coding parameters of an input coded image signal, and outputs the generated coding parameter to the decoded image signal. The image signal processing system according to claim 6, wherein the image signal is output together with the image signal. 8. The decoder outputs a picture type and / or a motion vector together with the decoded image signal as the coding parameter, in addition to a coding bit amount and / or an average quantization scale. The image signal processing system according to claim 7. 9. The image signal processing system according to claim 7, wherein said decoder outputs, together with said decoded image signal, an encoding parameter used when performing image encoding on a picture basis. 10. The image signal processing system according to claim 7, wherein the decoder outputs, in units of slices, an encoding parameter used when performing image encoding together with the decoded image signal. 11. The image signal processing system according to claim 7, wherein said encoder performs rate control based on an input coded bit amount and / or an average quantization scale. 12. A coding parameter generating means for generating a coding bit amount and / or an average quantization scale as a representative value of a coding parameter of an input coded image signal; Decoding means for decoding the image data to generate a decoded image signal, and outputting the encoding parameter generated by the encoding parameter generating means together with the decoded image signal generated by the decoding means. 13. The coding parameter generating means generates a picture type and / or a motion vector in addition to a coding bit amount and / or an average quantization scale as a representative value of the coding parameter. The decoder according to claim 12, wherein 14. The decoder according to claim 12, wherein the encoding parameter generated by said encoding parameter generating means is used when performing image encoding on a picture-by-picture basis. 15. The decoder according to claim 12, wherein the encoding parameters generated by said encoding parameter generating means are used when performing image encoding in units of slices. 16. A decoding unit for decoding an input encoded image signal to generate a decoded image signal, and corresponding to an invalid period in a signal format for transmitting the decoded image signal generated by the decoding unit. Output means for superimposing an encoding parameter of the input encoded image signal on a signal portion and outputting the encoding parameter together with the decoded image signal. 17. An image processing apparatus comprising: an encoding parameter generation unit configured to generate an encoding bit amount and / or an average quantization scale as a representative value of an encoding parameter of an input encoded image signal; 17. The decoder according to claim 16, wherein the encoding parameter generated by the encoding parameter generating means is output together with the decoded image signal. 18. The method according to claim 18, wherein said coding parameter generating means generates a picture type and / or a motion vector in addition to a coding bit amount and / or an average quantization scale as said coding parameter. Item 17
A decoder as described. 19. The image processing apparatus according to claim 17, wherein said encoding parameter generated by said encoding parameter generating means is an encoding parameter used when performing image encoding on a picture basis. Decoder. 20. The image processing apparatus according to claim 17, wherein said coding parameter generated by said coding parameter generating means is a coding parameter used when performing image coding on a slice basis. Decoder. 21. An encoded bit amount and / or an average quantization scale is generated as a representative value of an encoding parameter of an input encoded image signal, and the input encoded image signal is decoded to decode a decoded image. An image signal processing method, comprising: generating a signal; outputting the encoding parameter together with the generated decoded image signal; and performing image encoding on the decoded image signal using the encoding parameter. 22. A decoded image signal is generated by decoding the input coded image signal, and the signal portion corresponding to an invalid period in a signal format for transmitting the generated decoded image signal is provided. Superimposing the encoding parameter of the encoded image signal, outputting the encoding parameter together with the decoded image signal, and performing image encoding on the decoded image signal using the encoding parameter. Image signal processing method. 23. A coded bit amount and / or an average quantization scale is generated as a representative value of a coding parameter of an input coded image signal, and the input coded image signal is decoded to obtain a decoded image. A decoding method, comprising: generating a signal; and outputting the encoding parameter together with the decoded image signal. 24. An input coded image signal is decoded to generate a decoded image signal, and the input decoded signal is transmitted to a signal portion corresponding to an invalid period in a signal format for transmitting the generated decoded image signal. And decoding the encoded image signal together with the decoded image signal and outputting the encoded parameter together with the decoded image signal.