JP2002152047A - Multi-program compression coding signal conversion method and device, and medium with conversion program recorded thereon - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a processing apparatus having a bit rate conversion func tion that aims at a MPEG-2 transport stream obtained by multiplexing multiple programs. SOLUTION: A demultiplexer demultiplexes a received MPEG-2 TS into video TS packets by the programs, trans-coding is applied to each program, information denoting an input decoding quantity and an output coding quantity generated by each program is collectively controlled by rate control so as to enhance the overall image quality obtained from the stream.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a multi-program compression / encoding signal conversion method and apparatus, and a medium recording a conversion program, and more particularly to a multiplexed multimedia stream (MPEG) including a plurality of pieces of data such as program information.
TECHNICAL FIELD The present invention relates to a multi-program compression / encoding signal conversion method and apparatus for performing a bit rate conversion process on a (-2) multi-program transport stream or simply a transport stream) and a medium on which a conversion program is recorded.

[0002]

2. Description of the Related Art In a technique for digitizing a moving image, as a method for compressing and encoding a huge amount of information generated, a standard ISO / IEC 13 standard for an encoding method for digital video and accompanying audio is used.
818 (commonly known as “MPEG-2” (Moving Pi)
culture Expert Group Phase
2)). MPEG-2 generated in this manner
Bit stream (hereinafter referred to as “MPEG
-2 bit stream ") is used in a wide range of fields such as communication and television broadcasting.

[0003] The MPEG-2 bit stream has a hierarchical structure, and a GOP (Group) starts from the highest sequence layer.
of Pictures) layer, a picture layer, a slice layer, a macroblock layer, and a block layer.

In MPEG-2, in a moving image composed of a series of a plurality of screens, each screen is temporarily stored in a frame memory, and a difference between frames is calculated to reduce redundancy in a time axis direction. Further, a plurality of pixels constituting each frame are subjected to discrete cosine transform (hereinafter, “DC
T ") to reduce the redundancy in the spatial axis direction, thereby realizing efficient moving image compression encoding.

[0005] The coded signal is sent to a decoder and decoded and reproduced. In the decoder, the screen is reproduced and stored in the first frame memory, the next screen to be predicted is predicted based on the difference information, and stored in the second frame memory. Further, by making a prediction, a series of screens are formed and a moving image is reproduced. Such an approach is called bidirectional prediction.

In MPEG-2, three types of I picture, P picture and B picture are defined in order to realize the bidirectional prediction. The I picture is an abbreviation of an intra-coded picture, and is a screen that is coded as a still image independently of other pictures. P
A picture is an abbreviation of a forward prediction coded picture, and is a screen that is predictively coded based on an I or P picture located in the past in time. B picture is an abbreviation of bidirectional predictive coded picture, and is a picture that is predictively coded based on a forward, backward or bidirectional picture using I or P pictures located before and after in time. It is. That is, after the I picture and the P picture are coded first, the B picture inserted between them is coded.

[0007] The MPEG-2 bit stream encoded by the encoder is transmitted to a transmission line at a predetermined transfer rate.
The data is input to a decoder on the transmission path, decoded and reproduced.
However, the amount of information generated by encoding a moving image is not constant. Particularly, at the time of a scene change, the amount of information increases at a stretch. In order to send such a non-constant coded signal to a fixed-rate transmission path, it is necessary to control the rate of coded data in advance so that the amount of information that exceeds the level of the transmission buffer does not occur.

In MPEG-2, ISO / IEC JT
C1 / SC29 / WG11 / N0400 Test M
model 5 (April, 1993)
The rate control method is described in “TM5”.

[0009] In the rate control of TM5 of MPEG-2,
In step 1, first, bits are allocated for each picture type based on the code amount R allocated to an uncoded picture in a GOP. In step 2, a quantization scale used when performing the encoding process in units of macroblocks is calculated from the virtual buffer occupancy calculated based on the bit allocation.

Also, since there are a large number of decoders having a compression format other than MPEG-2 and decoders connected to transmission lines having different transmission speeds, MPEG-2 bitstreams having different compression formats and different transmission speeds are used. A moving image compression / encoding signal converter for conversion is required. A device for realizing this is a so-called transcoder.
The image compression encoded signal transmitted from the encoder is converted into an appropriate signal by a transcoder, and the signal is supplied to each decoder.

FIG. 20 shows a general conventional transcoder 5.
0 shows a first example. The conventional transcoder 50 has a first
A variable length decoding unit (denoted as separation / VLD) 51 connected to a first transmission line (not shown) having a bit rate and receiving the first MPEG-2 bit stream b1, an inverse quantizer 53, Unit 55 and a second transmission path (not shown) having a second bit rate, and a second MPEG
VLC 57 that outputs -2 bit stream b2, and rate controller 5 that controls the amount of code generated by quantizer 55
9 is provided. The second bit rate is a transfer rate lower than the first bit rate.

The VLD 51 and the inverse quantizer 53 decode the first MPEG-2 bit stream b1 up to the DCT coefficient area in macroblock units, and
And the VLC 57 encodes the obtained DCT coefficient signal to generate a second MPEG-2 bit stream b2 having a smaller code amount than the first MPEG-2 bit stream.

In the quantization process in the quantizer 55, D
The coefficient obtained by the CT transformation is divided by a predetermined quantization step. As a result, the image signal is compressed. This quantization step is obtained by multiplying a plurality of quantization matrix values included in a predetermined quantization table by a quantization scale.

In the transcoder 50, the first MPEG-
A sequence layer, a GOP layer in the two-bit stream b1,
Almost the coding information of the picture layer, slice layer and macroblock layer is reused. Basically DCT of block layer
Only the process of converting the code of the macro block layer which needs to be modified in accordance with the conversion of the coefficient and the conversion of the block layer is performed.

The transcoder 50 constructed as described above
In, the rate control section 59 performs rate control described in TM5 of MPEG-2. FIG. 21 shows a flowchart of the rate control process of the conventional transcoder 50. As shown in the figure, the conventional rate control process includes steps A1 to A14.

In step A1, a variable n is set to 1.
Here, the variable n indicates a number assigned to a plurality of pictures included in the input image signal, and hereinafter the n-th picture is indicated as pic (n).

In step A2, indices Xi, Xp and Xb indicating the complexity of the I, P and B pictures are calculated by the following equations (a1), (a2) and (a3).

Xi = Si × Qi Expression (a1) Xp = Sp × Qp Expression (a2) Xb = Sb × Qb Expression (a3)

Here, Si, Sp and Sb are generated code amounts of I, P and B pictures, respectively, and Qi, Qp
And Qb are average quantization parameters which are the average values of the quantization scale codes of all the macroblocks in the I, P and B pictures, respectively. However, the average quantization parameter is normalized in the range of 1 to 31.

This screen complexity index Xi, Xp and Xb
Is large for an image in which a large amount of encoded information is generated, that is, an image with a low compression rate, and is small for an image with a high compression rate.

The initial values of the parameters Xi, Xp and Xb indicating the complexity of the screen of the I, P and B pictures are as follows:
It is given by the following equations (a4), (a5) and (a6).

Xi = 160 × target_Bitrate / 115 Expression (a4) Xp = 60 × target_Bitrate / 115 Expression (a5) Xb = 42 × target_Bitrate / 115 Expression (a6)

Here, target_Bitrate is a target bit rate of the transcoder 50.

In the following step A3, the code amounts Ti, Tp and Tb to be allocated to the I, P and B pictures in the GOP are calculated by the following equations (a7), (a8) and (a9). Here, Np and Nb indicate the numbers of uncoded P and B pictures in the GOP, respectively.

(Equation 1) Here, Kp and Kb indicate the ratios of the quantized scale codes of the P and B pictures with respect to the quantized scale codes of the I picture, and Kp = 1.0 and Kb = 1.
Assume that the overall image quality is always optimized when it becomes 4.

In the following step A4, it is determined whether or not the variable n is 1. That is, the picture to be coded is 1
It is determined whether or not the picture is pic (1). If it is the first picture, the process proceeds to step A5. If it is not the first picture, the process proceeds to step A6. In step A5, the code amount R allocated to the uncoded picture in the GOP when the first picture pic (1) in the GOP is coded is calculated by the following equation (a10).

R = target_Bitrate × N / picture_rate + R Expression (a10)

Where N is the total number of pictures in the GOP,
picture_rate is a value indicating the time resolution of the input image, and indicates the number of screens decoded and displayed in one second.

In step A6, the code amount R to be allocated to the uncoded picture in the GOP is changed to the generated codes of the I, P and B pictures when the (n-1) th picture pic (n-1) is coded. Based on the quantity Si, Sp, or Sb, updating is performed by one of the following equations (a11), (a12), and (a13).

R = R-Si formula (a11) R = R-Sp formula (a12) R = R-Sb formula (a13)

Steps A5 and A6 both proceed to step A7 to set 1 to a variable j. Where the variable j
Indicates a number assigned to a plurality of macroblocks in one picture, and hereinafter the j-th macroblock is indicated as MB (j).

In the following step A8, the occupation amounts di (j), dp (j) and db (j) of the virtual buffer when encoding the j-th macroblock MB (j) in the I, P and B pictures. Are calculated by the following equations (a14), (a15) and (a16).

(Equation 2) Here, B (j-1) is the (j-1) th macroblock M
This is the generated code amount of all macroblocks up to B (j-1).

Further, di (0), dp (0) and db (0) are initial values of the virtual buffer occupancy of the I, P and B pictures, respectively, and are given by the following equations (a17), (a18) and Each is given by equation (a19).

Di (0) = 10 × r / 31 Equation (a17) dp (0) = Kp × di (0) Equation (a18) db (0) = Kb × di (0) Equation (a19)

Here, r is called a reaction parameter, and is represented by the following equation (a20), and controls the response speed of the feedback loop.

R = 2 × target_Bitrate / picture_rate Equation (a20)

The virtual buffer occupancy at the end of I, P, and B picture encoding, that is, the virtual buffer occupancy di (NMB), dp (NMB) when the NMB-th macroblock MB (NMB) is encoded. And db (NMB) are used as initial values di (0), dp (0) and db (0) of the virtual buffer occupancy at the next encoding for each picture type.

In the following step A9, the quantization scale code Q (j) for the j-th macroblock MB (j) for each picture based on the virtual buffer occupancy d (j).
Is obtained by the following equation (a21).

Q (j) = d (j) × 31 / r Equation (a21)

In the following step A10, the j-th macroblock MB (j) is quantized using the quantization scale code Q (j) calculated in step A9. Subsequent step A
In step 11, the variable j is incremented, and step A12
Then, it is determined whether or not the variable j exceeds the total number of macroblocks NMB. Here, NMB is the total number of macroblocks included in the n-th picture pic (n).
If the variable j does not exceed the total number of macroblocks NMB, the process returns to step A8. If the variable j exceeds the total number of macroblocks NMB, the process proceeds to step A13.

As described above, the variable j is set in step A8.
It is also used as a loop counter for repeating the encoding process of A11. Thereby, the n-th picture pi
Encoding processing can be sequentially performed on all macroblocks from the first macroblock MB (1) to the NMBth macroblock MB (NMB) in c (n).

In step A13, the variable n is incremented, and the flow advances to step A14 to determine whether or not the variable n exceeds the total number NPIC of pictures to be encoded. If the variable n does not exceed the total number of pictures NPIC, the process returns to step A2. If the variable n exceeds the total number of pictures NPIC, the process ends.

As described above, since the first transcoder 50 cannot have information about the image structure such as the I and P picture periods, the TM5 shown in FIG.
A method of allocating bits based on information such as an image GOP structure, such as the rate control described above, cannot be performed unless an input image structure is assumed.

Therefore, there is a second conventional transcoder 60 shown in FIG. 22 as an example employing a method of performing rate control without assuming a GOP structure. As shown in the figure, a second conventional transcoder 60 includes a delay circuit 61, a bit rate ratio calculator 63, and an input code in addition to the configuration of the first conventional transcoder 50 of FIG. Amount integrator 65, difference code amount calculator 67, target output code amount updater 69, quantization scale code calculator 71
And

The transcoder 60 thus configured
23 is shown in FIG. As shown in the figure,
The process of the transcoder 60 includes steps B1 to B13. Steps B6 to B13 are the same as steps A7 to A14 of the rate processing shown in FIG. However, in step B7, the virtual buffer occupancy is calculated based on the target output code amount Tout calculated by the target output code amount update unit 69.

FIGS. 24 and 25 show a third example of a conventional transcoder as another example in which the rate control is similarly performed without assuming the GOP structure. FIG.
As shown in FIG. 4, a third conventional transcoder 80
Is connected to a first transmission line having a first bit rate,
A VLD 81 for inputting an input bit stream b3 and an inverse quantizer 53, which is the same as the first conventional transcoder 50,
, A quantizer 55, and a VLC 57, and the same bit rate ratio calculation unit 63 as the transcoder 60 in FIG.
, A difference code amount calculation unit 67, a target output code amount update unit 83, and a quantization scale code calculation unit 8
5 is provided.

In the third conventional transcoder 80, the code amount is described in advance in the bit stream b3 as information, and rate control is performed based on the information.

However, since the transcoder is intended for the signal after the encoding process, it cannot know the original signal before the encoding. Therefore, in the code amount control, attention is paid not to the distortion of the image itself after the transcoding process, but to the distortion newly generated by the requantization process, and by suppressing this distortion, the degradation of the image quality is suppressed. The amount of code must be reduced.

Accordingly, the applicant of the present application has previously filed Japanese Patent Application No.
No. 278867 and Japanese Patent Application No. 11-327384 were filed.

The technique described in Japanese Patent Application No. 11-278867 discloses a decoding quantization parameter and a re-quantization rate distortion function which depends on a decoding quantization parameter and a re-quantization parameter. A moving image compression / encoding signal conversion method and apparatus for realizing calculation of an optimal quantization parameter based on a calculated quantization parameter, and a medium recording a conversion program.

In this transcoder, an inverse quantizer for performing inverse quantization and a quantizer for performing requantization are quantized by considering a rate distortion function based on an input quantization parameter. By providing the quantization parameter switching unit that switches the parameters, it is possible to minimize the error when converting the quantized coefficient area data to the requantized coefficient area data.

Also, the one described in Japanese Patent Application No. 11-327384 describes the size of the quantization parameter obtained from the input bit stream by taking into account the reduced code amount and the generated distortion in the requantization processing in the transcoder. A moving image compression coded signal conversion method, apparatus, and a medium on which a conversion program is realized to realize a transcoder code amount control method for controlling a reduced code amount according to the amount and minimizing a distortion generated due to requantization. is there.

This is a target reduced code amount calculator that calculates an average reduced code amount that is a target reduced code amount for each input quantization parameter value, an average reduced code amount of the input quantization parameter value of the region, and A target code amount calculator that calculates a target code amount based on the code amount of the orthogonal transform coefficient region data of the region, and a quantization parameter based on the target code amount calculated by the target code amount calculator. By providing a quantization scale code calculator to be set, it is possible to minimize distortion generated in the requantization processing in the transcoder.

[0053]

SUMMARY OF THE INVENTION However, MPE
The G-2TS has a syntax that can support multi-programs in order to realize the broadcasting of multiple programs. For various video distribution services such as BS digital broadcasting and CATV,
It is assumed that a multi-program transport stream obtained by multiplexing a plurality of programs and the like (hereinafter, referred to as a multi-program TS) is used.

Therefore, when an actual content delivery service is operated, it is difficult to think of a content consisting of only a single video bit stream, and the service is provided as a multiplexed multimedia stream (multi-program TS) in which a plurality of programs are multiplexed. It is thought that there are many opportunities to be done.

However, at present, almost no studies have been made on transcoders compatible with the MPEG-2 multi-program TS. In particular, multi-program T
In the transcoding process for S, the video elementary stream (video E
It is necessary to dynamically fluctuate based on the coding characteristic of S). Furthermore, since the picture to be transcoded is not perfect due to the formation of a transport stream packet (hereinafter, referred to as a TS packet), it is necessary to establish a control method in consideration of the asynchronousness of the picture.

The following describes the MPEG-2 multiprogram T
The requirements for implementing the code amount control method in transcoding for S will be summarized.

-Difference in video ES rate (Issue 1)

Even when the video ESs constituting each program have the same sequence, if the coding rates of the video ESs are different, and if the information amount is reduced (transcoded) at the same code amount reduction rate, The generated error largely depends on the ES coding rate.

When each video ES is encoded by VBR, each ES encoding rate fluctuates as needed.

Further, since the video ES packetization depends on the TS multiplexer, the coding rate of the video ES included in the TS bit stream to be transcoded and the number of packets are not necessarily proportional.

As described above, a control method that takes into account not only the number of input packets and the input code amount but also the encoding rate of the video ES is required.

-Difference in video ES coding characteristics (Problem 2)

Even when the video ESs have the same coding rate, the coding characteristics of each sequence are greatly different. Therefore, even in the case of the same rate, control considering the coding efficiency is required.

Further, since the coding characteristics always fluctuate even when there is no scene change, a dynamic control method taking the coding characteristics into consideration is required.

Asynchrony of picture (Issue 3)

Since there is no synchronization of the video ES between the programs, the picture configuration in each program included in the transcoding target TS is different.

Further, T within a unit time (constant observation time)
Since it is impossible to assume that the timing (start and end) of the S packet and the constituent picture match, the remaining code of the last picture of the previous unit time may exist at the start of the unit time. For this picture, when the target code amount is newly calculated, the requantization characteristic in the previous unit time greatly changes, and there is a possibility that the image quality may greatly change in the middle of the picture.

As described above, a code amount control method for suppressing a change in image quality in the middle of a picture is required.

Therefore, the present invention proposes a code amount control method for MPEG-2TS multi-program transcoding. Specifically, based on the picture configuration in the video ES existing during the buffered observation time (hereinafter, referred to as unit time bf_time), the rate distortion characteristics for each picture type, and the picture asynchronism, We propose a code amount control algorithm that dynamically calculates the optimal target code amount of.

In the present invention, since the synchronization of the coded frames, which has been premised especially in the case where the conventional multi-encoder has been studied, is not guaranteed, the processing of pictures extending over each unit time is also taken into consideration. . Further, in the present invention, it is possible to cope with a change in image characteristics by calculating a rate distortion characteristic function as needed at the time of transcoding.

[0071]

According to the first aspect of the present invention,
In order to solve the above problem, an input step of inputting a first coded signal obtained by multiplexing a plurality of programs and performing compression coding via a first transmission path having a first transfer rate;
A signal demultiplexing step of demultiplexing the first encoded signal input in the input step into a first data sequence, a second data sequence, and a third data sequence; and a plurality of first demultiplexed signals separated in the signal demultiplexing step. A data string converting step of generating a plurality of converted first data strings having a smaller total code amount than a total code amount of the plurality of first data strings from the data string; and a third data string separated in the signal demultiplexing step. Is corrected in accordance with the characteristic change after the conversion of the first coded signal to generate a corrected third data sequence, and the converted first data sequence generated in the data sequence conversion step is ,
A signal array multiplexing step of multiplexing the second data string separated in the signal demultiplexing step and the corrected third data string corrected in the third data string correcting step to generate a second encoded signal And an output step of outputting the second encoded signal via a second transmission path having a second transfer rate lower than the first transfer rate,

In the data string conversion step, when the code amount conversion of the first data string to the converted first data string is performed,
The code amount conversion of the first data string is performed for each program, and information of another program is also acquired to determine the code amount conversion amount.

According to a second aspect of the present invention, there is provided a multi-program compression-encoded signal conversion method according to the first aspect, wherein the inputting step comprises:
An MPEG-2 transport stream is input as an encoded signal, and the signal demultiplexing step separates a transport stream packet including a compression-encoded video signal as the first data stream. It is characterized by outputting an MPEG-2 transport stream as a second encoded signal.

According to a third aspect of the present invention, there is provided a multi-program compression-encoded signal conversion method according to the first aspect, wherein the inputting step comprises:
A multiplexed audio / video compression coded stream is input as a coded signal, and the signal demultiplexing step separates a video partial stream as the first data stream, and the output step outputs the second coded signal as the second coded signal. It is characterized by outputting a multiplexed audio / video compression encoded stream.

According to a fourth aspect of the present invention, there is provided a multi-program compression-encoded signal conversion method according to the second or third aspect, wherein the data string conversion step comprises the step of: A requantization parameter serving as a reference of a conversion amount of the code amount conversion is set for each information, and the image information is code amount converted according to the requantization parameter.

According to a fifth aspect of the present invention, there is provided a multi-program compression coded signal conversion method according to the fourth aspect, wherein the second coded signal is generated from the first coded signal. All the data string conversion processing for the predetermined unit time interval as a processing unit, the data string conversion step inputs the first data string within the unit time interval, The code amount conversion amount of the first data string is determined based on information of one data string.

According to a sixth aspect of the present invention, there is provided a multi-program compression-encoded signal conversion method according to the fifth aspect, wherein the data string conversion step comprises:
When one piece of image information of the first data string is input over the unit time, the remaining image information part of the converted image information is converted into the converted image information in the unit time. The code amount conversion is performed according to the requantization parameter used when the image information is converted.

According to a seventh aspect of the present invention, there is provided a multi-program compression-encoded signal conversion method according to any one of the fourth to sixth aspects, wherein the data string conversion step comprises the steps of: Image information to be converted
A virtual information storage amount is calculated based on the code amount before the image information input within the unit time and the converted code amount of the image information before the image information, and the virtual information storage amount is calculated based on the virtual information storage amount. It is characterized by setting a quantization parameter.

According to an eighth aspect of the present invention, there is provided a multi-program compression-encoding signal conversion method according to the seventh aspect, wherein the data string conversion step comprises:
The virtual information storage amount is determined based on an input / output code amount ratio which is a ratio between an input decoded code amount obtained by decoding the input code amount of the image information and a target target code amount of the code amount conversion amount of the image information. Is calculated.

According to a ninth aspect of the present invention, in order to solve the above-mentioned problem, in the multi-program compression-encoded signal conversion method according to any one of the fourth to eighth aspects, the data string conversion step includes the step of: The method is characterized in that the requantization parameter serving as a reference when converting the image information of the first data string into a code amount is set based on the configuration information of the image.

According to a tenth aspect of the present invention, in order to solve the above problems, in the multi-program compression-encoded signal conversion method according to any one of the fourth to ninth aspects, the data sequence conversion step The first image information obtained by decoding the image information of the first data sequence
A quantization error generated from the quantization parameter and a reference quantization parameter serving as a reference of the code amount conversion calculated by calculation is smaller than a quantization error generated in the code amount conversion process using the reference quantization parameter. The re-quantization parameter is set.

According to an eleventh aspect of the present invention, in order to solve the above-mentioned problem, a plurality of programs are multiplexed and a first encoded signal, which has been compression-encoded, is transmitted via a first transmission line having a first transfer rate. An input means for inputting the first coded signal, a signal demultiplexing means for separating the first coded signal input by the input means into a first data sequence, a second data sequence and a third data sequence; A data string converting means for generating a plurality of converted first data strings having a total code amount smaller than a total code amount of the plurality of first data strings from the plurality of separated first data strings; A third data string correcting unit that corrects the separated third data string according to the characteristic change after the conversion of the first encoded signal to generate a corrected third data string, and a third data string generated by the data string converting unit. The converted first data sequence and the signal multiplex Signal array multiplexing means for multiplexing the second data string separated by the separation means and the corrected third data string corrected by the third data string correction means to generate a second encoded signal; Output means for outputting the second encoded signal via a second transmission path having a second transfer rate lower than the first transfer rate,

When the data string conversion means performs the code amount conversion of the first data string to the converted first data string, the data string conversion means performs the code amount conversion of the first data string for each program, and also converts information of another program. The code amount conversion amount is also determined by acquiring the code amount conversion amount.

According to a twelfth aspect of the present invention, in order to solve the above-mentioned problems, in the multi-program compression-encoded signal conversion apparatus according to the eleventh aspect, the input means comprises the first program.
An MPEG-2 transport stream is input as an encoded signal, the signal demultiplexing unit separates a transport stream packet including a compression-encoded video signal as the first data stream, and the output unit outputs It is characterized by outputting an MPEG-2 transport stream as a second encoded signal.

According to a thirteenth aspect of the present invention, there is provided a multi-program compression-encoded signal converting apparatus according to the eleventh aspect, wherein the input means comprises:
A multiplexed audio / video compression-coded stream is input as a coded signal, the signal demultiplexing unit separates a video partial stream as the first data stream, and the output unit outputs the second coded signal as the second coded signal. It is characterized by outputting a multiplexed audio / video compression encoded stream.

According to a fourteenth aspect of the present invention, in the multi-program compression coded signal conversion device according to the twelfth or thirteenth aspect, the data sequence conversion means comprises an image of the first data sequence. A requantization parameter serving as a reference of a conversion amount of the code amount conversion is set for each information, and the image information is code amount converted according to the requantization parameter.

According to a fifteenth aspect of the present invention, there is provided a multiprogram compression-encoded signal conversion apparatus according to the fourteenth aspect, wherein the second encoded signal is generated from the first encoded signal. All the data string conversion processing for the predetermined unit time interval as a processing unit, the data string conversion means inputs the first data string within the unit time interval, The code amount conversion amount of the first data string is determined based on information of one data string.

According to a sixteenth aspect of the present invention, in order to solve the above-mentioned problem, in the multi-program compression-encoded signal converting apparatus according to the fifteenth aspect, the data string converting means comprises:
When one piece of image information of the first data string is input over the unit time, the remaining image information part of the converted image information is converted into the converted image information in the unit time. The code amount conversion is performed according to the requantization parameter used when the image information is converted.

According to a seventeenth aspect of the present invention, there is provided a multi-program compression-encoding signal conversion apparatus according to any one of the fourteenth to sixteenth aspects, wherein
The data sequence conversion means, the image information of the code amount conversion, the code amount before the image information input in a unit time,
A virtual information storage amount is calculated based on the converted code amount of the image information before the image information, and the requantization parameter is set based on the virtual information storage amount.

According to an eighteenth aspect of the present invention, in order to solve the above-mentioned problems, in the multi-program compression-encoded signal converting apparatus according to the seventeenth aspect, the data sequence converting means comprises:
The virtual information storage amount is determined based on an input / output code amount ratio which is a ratio between an input decoded code amount obtained by decoding the input code amount of the image information and a target target code amount of the code amount conversion amount of the image information. Is calculated.

According to a nineteenth aspect of the present invention, there is provided a multi-program compression-encoding signal conversion apparatus according to any one of the fourteenth to eighteenth aspects, wherein
The data sequence conversion means sets the requantization parameter serving as a reference when converting the image information of the first data sequence into a code amount based on the configuration information of the image. .

According to a twentieth aspect of the present invention, there is provided a multi-program compression coded signal conversion apparatus according to any one of the fourteenth to nineteenth aspects, wherein
The data sequence converting means encodes the image information obtained by decoding the image information of the first data sequence.
A quantization error generated from the quantization parameter and a reference quantization parameter serving as a reference of the code amount conversion calculated by calculation is smaller than a quantization error generated in the code amount conversion process using the reference quantization parameter. The re-quantization parameter is set.

According to a twenty-first aspect of the present invention, in order to solve the above-mentioned problem, a plurality of programs are multiplexed and a first coded signal which is coded and compressed is transmitted via a first transmission line having a first transfer rate. An input step of inputting the first encoded signal, a signal demultiplexing step of demultiplexing the first coded signal input in the input step into a first data sequence, a second data sequence, and a third data sequence. Separated first plurality
A data string converting step of generating a plurality of converted first data strings having a smaller total code amount than a total code amount of the plurality of first data strings from the data string; and a third data string separated in the signal demultiplexing step. Is corrected in accordance with the characteristic change after the conversion of the first coded signal to generate a corrected third data sequence, and the converted first data sequence generated in the data sequence conversion step is Multiplexing the second data string separated in the signal demultiplexing step and the corrected third data string corrected in the third data string correcting step to generate a second encoded signal. And outputting the second encoded signal via a second transmission path having a second transfer rate lower than the first transfer rate.

In the data string conversion step, when the code amount conversion of the first data string to the converted first data string is performed,
The code amount conversion of the first data string is performed for each program, and information of another program is also acquired to determine the code amount conversion amount.

According to a twenty-second aspect of the present invention, in order to solve the above-mentioned problem, in the medium recording the multi-program compression-encoded signal conversion program according to the twenty-first aspect, the inputting step comprises the steps of: MPEG-
2 transport stream, the signal demultiplexing step separates a transport stream packet including a compression-encoded video signal as the first data stream, and the output step performs the second encoded signal as the second encoded signal. It is characterized by outputting an MPEG-2 transport stream.

According to a twenty-third aspect of the present invention, in order to solve the above-mentioned problem, in the medium recording the multi-program compression-encoded signal conversion program according to the twenty-first aspect, the inputting step comprises the step of: A multiplexed audio / video compression coded stream is input, the signal demultiplexing step separates a video partial stream as the first data stream, and the output step outputs a multiplexed audio / video as the second coded signal. It is characterized by outputting a moving picture compression encoded stream.

According to a twenty-fourth aspect of the present invention, in order to solve the above-mentioned problems, in the medium having recorded thereon the multi-program compression-encoded signal conversion program according to the twenty-second or twenty-third aspect, the data string conversion step includes the step of: For each image information of a data string, a requantization parameter serving as a reference of the conversion amount of the code amount conversion is set, and the image information is code amount converted according to the requantization parameter. is there.

According to a twenty-fifth aspect of the present invention, in order to solve the above-mentioned problem, in a medium recording a multi-program compression-encoded signal conversion program according to the twenty-fourth aspect, the second encoded signal is converted from the first encoded signal to the second encoded signal. All data string conversion processing for generating a signal is performed with a predetermined unit time interval as a processing unit, and the data string conversion step inputs the first data string within the unit time interval,
A code amount conversion amount of the first data string is determined based on information of the first data string in the unit time.

According to a twenty-sixth aspect of the present invention, in order to solve the above-mentioned problem, in the medium storing the multi-program compression-encoded signal conversion program according to the twenty-fifth aspect, the data string conversion step includes the step of: When one piece of image information is input across the unit time, the remaining image information part of the converted image information is converted into the converted image information in the unit time when the remaining image information part is input. The code amount conversion is performed according to the requantization parameter used at the time.

According to a twenty-seventh aspect of the present invention, in order to solve the above-mentioned problems, the data string conversion method according to any one of the twenty-fourth to twenty-sixth aspects is provided for a medium on which the multi-program compression-encoded signal conversion program is recorded. The step of converting the image information to be subjected to code amount conversion into a virtual information storage amount based on the code amount before the image information input within a unit time and the converted code amount of the image information before the image information. Is calculated, and the requantization parameter is set based on the virtual information accumulation amount.

According to a twenty-eighth aspect of the present invention, in order to solve the above-mentioned problems, in the medium recording the multi-program compression-encoded signal conversion program according to the twenty-seventh aspect, the data string conversion step includes the step of inputting the image information. Calculating the virtual information storage amount based on an input / output code amount ratio which is a ratio between an input decoded code amount obtained by decoding the code amount and a target target code amount of the code amount conversion amount of the image information. It is a feature.

According to a twenty-ninth aspect of the present invention, in order to solve the above-mentioned problem, the medium for recording the multi-program compression-encoded signal conversion program according to any one of the twenty-fourth to twenty-eighth aspects is characterized in that: The step is characterized in that the re-quantization parameter, which is a reference when converting the image information of the first data string into a code amount, is set based on the configuration information of the image.

According to a thirtieth aspect of the present invention, in order to solve the above-mentioned problems, the data sequence conversion method according to any one of the twenty-fourth to twenty-ninth aspects is provided on a medium on which the multi-program compression-encoding signal conversion program is recorded. A step of: coding a first quantization parameter when the image information obtained by decoding the image information of the first data string is encoded; and a reference quantization parameter serving as a reference for code amount conversion calculated by calculation. The re-quantization parameter is set so that a quantization error generated by the code amount conversion process using the reference quantization parameter is smaller than a quantization error generated by the reference quantization parameter.

[0104]

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

FIG. 1 shows a rate converter representing an outline of the present invention.

This rate converter 600 is an MPEG-2T
S demultiplexer 610, MPEG-2TS multiplexer 62
0, an MPEG-2 video transcoder 640 and a system controller 650.

The MPEG-2 TS demultiplexer 610 separates an input MPEG-2 transport stream into a video TS (transport stream), an audio TS, and system information TS, and outputs each.

The MPEG-2 TS multiplexer 620 receives and multiplexes the video TS, the audio TS, and the system information TS, and outputs the resultant as an output MPEG-2 transport stream.

MPEG-2 Video Transcoder 640
Is for inputting a video TS, performing transcoding (coding amount conversion) processing, and outputting a transcoded video TS.

Here, the input video TS includes a plurality of programs. Therefore, the video elementary stream (ES) decoded from the video TS is
Transcoding is performed for each video ES constituting each program. At this time, the code amount of each video ES is reduced for each program. However, for the encoding rate control, the encoding process is performed in consideration of the code amount of another video ES (stream of another program). . Details will be described later.

The system controller 650 has the system information T
MPE which inputs S and outputs from this rate converter 600
It converts the system information TS to the G-2 transport stream and outputs the converted system information TS.

The rate converter 600 considers a video bit stream, which is considered to have a relatively large bit rate occupation in the MPEG-2 transport stream, as an object of bit rate reduction, and separates the video bit stream by the MPEG-2 TS demultiplexer 610. Only the video TS that was
The bit rate is reduced by the G-2 video transcoder 640, and the audio TS and the system information TS are subjected to the processing of changing the information of the input MPEG-2 transport stream as it is or a part of the fixed length code,
The system is simplified by using it as information of an output MPEG-2 transport stream.

Further, the rate converter 600 is an MPEG converter.
Problems due to the simple combination of the MPEG-2TS decoder and the MPEG-2TS encoder:-The processing amount is large. -Image quality is degraded due to repeated coding. -A delay occurs due to the rearrangement of frames. And the MPEs to be transcoded
1. of each program in the G-2 transport stream 1. Configuration of a plurality of pictures 2. Rate distortion characteristics for each picture type We implement a picture optimal target code amount derivation method and code amount control algorithm in consideration of picture asynchrony.

FIG. 2 is a block diagram showing a simplified visual representation of the features of the multiple program processing of the present invention. The portion surrounded by the dotted line is the video transcoder.

In the conventional processing, the rate control is performed for each program. In the present processing, as shown in FIG. 2, the code amount is reduced for each program. All programs are collectively processed according to the coding rate.

Before explaining the code amount control in a plurality of programs, a video TS and a non-video TS, that is, an audio T
The difference in processing between TSs that are not targeted for reduction, such as S, will be described.

First, a rate for ensuring the synchronization of the video bit stream by possessing PTS (Presentation Time Stamp: display time management information) and DTS (Decoding Time Stamp: decoding time management information) of the input MPEG-2 transport stream. The converter is shown in FIG.

As shown in FIG. 3, the present rate converter 700
Is an MPEG-2TS demultiplexer 710, an MPEG-2
TS multiplexer 720, video TS decoder 741, video PES decoder 742, video ES transcoder 74
4. Video PES packet generator 745 and video T
An S packet generator 746 is provided.

The MPEG-2 TS demultiplexer 710 separates a video TS from an input MPEG-2 transport stream and outputs the video TS.

[0120] The MPEG-2TS multiplexer 720 receives a video TS and outputs it as an output MPEG-2 transport stream.

The video TS decoder 741 receives and decodes a video TS and outputs a video PES.

The video PES decoder 742 outputs the video PES
Input S, decode it, PTS, DTS and video E
It outputs S (elementary stream).

The video ES transcoder 744 receives the video ES, performs transcoding, reduces the bit rate, and outputs the transcoded video ES.

[0124] The video PES packet generator 745
Input TS, DTS and video ES, and input video PE
S is generated and output.

The video TS packet generator 746 receives a video PES, generates a video TS, and outputs the video TS.

The present rate converter 700 is composed of PTS, DTS
Holds the PTS and DTS of each video elementary stream when decoding the PES packet, and outputs the MPEG
-2, the same video elementary stream of the input / output MPEG-2 transport stream and the same audio of the audio bit stream are encoded as PTS and DTS of the same video elementary stream of the transport stream. The frames have the same PTS, DTS,
Audio and video synchronization can be matched.

Next, the relationship between the video TS in the input / output MPEG-2 transport stream and the non-reduction target TS is shown in FIG. FIG. 4 is an example when the bit rate of the input MPEG-2 transport stream is reduced to １ ／.

The non-reduction target TS packets include a transport packet including a video bit stream in the input MPEG-2 transport stream and a TS packet (for example, P
AT, PMT) are all TS packets that are not to be reduced except for transport packets that include the transport packets.

As shown in the drawing, the arrangement position of the non-reduction target TS packets in the output MPEG-2 transport stream is determined by the input / output bit rate ratio of the arrangement position of the non-reduction target TS packets in the input MPEG-2 transport stream. By setting the position, the same non-reduction target TS packet arrives at the same time based on the time reference of the decoder.

Next, VBV due to the difference in the code amount reduction mode
Transition of occupancy of buffer (Video Buffering Verifier: virtual buffer for code generation amount control) and DTS (Decoding
FIG. 5 shows the difference between the Time Stamp (decoding time management information).

FIG. 5A shows an example in which the code amount of the P and B pictures is actively reduced without reducing the code amount of the I picture. FIG. , P, and B pictures are made constant. Note that the upper part of each figure in FIG. 5 is the video ES of the input bit stream, and the lower part is the video ES after transcoding.

In FIG. 5, B indicates the reception buffer size, B (n) ^ * indicates the VBV buffer occupancy immediately before the n-th encoded image is decoded, and B (n) indicates the n-th It shows the occupancy of the VBV buffer immediately after the encoded image is decoded and the code amount for one frame is removed from the buffer. The VBV buffer occupancy must always indicate a value between 0 and B. The inclination of the straight line indicates the bit rate.

In the case of FIG. 5 (a), when the upper input bit stream is reproduced, the waiting time of the reception buffer until decoding of the first frame is a value indicated by DTS, while the lower output bit The waiting time of the receiving buffer before decoding the first frame of the stream is DT
It takes the value indicated by S ^ *.

At this time, DTS ^ *> DTS, and (DT
(S ^ *-DTS) time delay occurs. When such a delay occurs, not only the difference between the input and the output but also the synchronization with the audio is lost. Conversely, if DTS is added as the decoding time of the output bit stream, all the codes constituting the frame may not have arrived at the DTS time at the MPEG-2TS decoder. You have to wait.

In FIG. 5B, DTS ^ * = DTS, and all the codes constituting the corresponding frame have arrived at the DTS time at the MPEG-2TS decoder.
The decoding of the corresponding frame can be started at S time.

Therefore, a rate control method is used in which the code amount for each frame of the input / output bit stream is proportional to the bit rate ratio.

Next, FIG. 6 is a conceptual diagram of a transcoding code amount control method for a VBR (Variable Bit Rate) coding format video bit stream. FIG.
Converts the input MPEG-2 transport stream to 1 /
This is an example of conversion to a bit rate of 2.

Here, in the code amount control of a plurality of programs, when transcoding a video elementary stream and converting it into a transcoded video elementary stream, if there is a picture that extends over each processing unit time, the processing is performed. The picture processing over the unit time is also considered.

The details of the code amount control method for a plurality of programs will be described later.

Since the MPEG-2 transport stream is composed of fixed-length packets of 188 bytes, the number of packets in a certain time interval is determined from the input bit rate by setting a certain time interval. Similarly, the output side sets the same time interval as the input side, so that the target number of output packets within a certain time interval is determined.

Next, it is possible to know how many video TSs and non-reduction target TSs are included in a unit time from information obtained by demultiplexing packets in a certain unit time (n). At this time, as described above, if the non-reduction target TS packets are excluded from the rate conversion target, the number of input non-reduction target TS packets and the output schedule in the unit time (n) are calculated based on the target number of output packets per unit time (n). The number of packets obtained by subtracting the number of PAT and PMT packets is the number of packets obtained as output video TS packets.

Therefore, the rate of the video elementary stream obtained by decoding the input video TS packets may be controlled by transcoding so that the target output video TS packet number is obtained.

However, when a video elementary stream obtained by transcoding a video bit stream is converted into a PES packet and a TS packet, an overhead is generated by adding a packet header. Cannot be predicted exactly. Therefore, the amount of overhead generated in PES packetization and TS packetization up to the past (n-1) time is calculated, and the output code amount of the output video elementary stream is determined by using this value.

Hereinafter, the above conditions are summarized and the MPE of the present invention is summarized.
The basic policy of the G-2TS transcoder algorithm is described, the processing configuration is shown, and the detailed algorithm is described.

Basic policy: The MPEG-2 transport stream (MPEG-2TS) is composed of an encoded signal represented by a video bit stream and an audio bit stream, and a system control signal.

The system control signal is a PMT related to a program.
A PAT (Program Association Table: program table) for designating a packet identifier of a TS packet for transmitting (described later), and a packet identifier and related information of the TS packet transmitted by each encoded signal constituting the program. PMT (Program Map) to be transmitted
Table: Program map table, program correspondence table,
For each program number, PID (Packet Identification) of a packet in which a stream of video, audio, additional data, etc. constituting the program is transmitted, which is 13-bit stream identification information and is an attribute of an individual stream of the packet Is shown). The PID of the PMT itself is specified by the PAT), a CAT (Conditional Access Table) for transmitting individual information, and a PI of a packet for transmitting decryption information for descrambling in descrambling in pay broadcasting.
D), and NIT (Network Information Table: physical information about the transmission path) that associates transmission path information such as a modulation frequency with a broadcast program.

[0147] Of these signals, deleting the system control signal in the bit rate reduction processing is considered to possibly hinder the actual service, and is therefore hardly considered as a reduction target. Therefore, a video bit stream or an audio bit stream that is an encoded signal is considered as a target of the bit rate reduction processing.

However, MP @ ML (Main Profile
at Main Level: Classifies the performance of the MPEG-2 decoder; generally, the profile specifies the functional classification (difference in syntax), and the level specifies the difference in amount (image size, etc.)) Has a coding rate of 4 [Mbps] to 15 [Mbps], whereas the bit rate of an audio bit stream is at most about 384 [Kbps].

Therefore, only the video bit stream that is considered to have a relatively large bit rate occupation in the MPEG-2 transport stream is subjected to the bit rate reduction processing.

Each process uses a unit time as a processing unit with bf_time [sec] as a unit time interval. The unit time (n) is the number of bits [bit] of the total input MPEG-2TS packet to the MPEG-2TS transcoder.
PEG-2 transport stream bit rate TS
It is calculated by dividing by B_in [bps]. Therefore, when the total number of input MPEG-2 TS packets All_TS_in satisfies the following expression (1), it is defined that the unit time (n) has elapsed.

(Equation 3) Here, 188 is a fixed packet length [byte / packet],
8 is the number of bits per byte [bit / byte].

However, the input MPE to this transcoder
When the G-2 transport stream is, for example, a bit stream of a drift like digital broadcasting,
No immediate transcoding, first PAT, PMT
Is detected as an initial value for calculating the unit time.

Next, FIG. 7 shows a conceptual diagram of a code amount control method for a plurality of programs.

As shown in FIG. 7, the present transcoder performs processing in units of bf_time as described above. That is, data for bf_time is scanned, and transcoding processing is realized therein. The larger the bf_time is, the more optimal rate control can be performed. However, the buffering period becomes long for that, and the real-time property is lost. Therefore, in the transcoder emphasizing real time, at most 200 [msec] is the time used as bf_time.

Here, the bf_time does not change between the receiving side and the transmitting side, but the amount of data contained in the same bf_time differs. In FIG. 7, ten packets exist before transcoding, but after transcoding, only five packets exist. This is because the line through which the stream is to be streamed after transcoding is narrower than the line before transcoding. In addition, in the three programs after transcoding, the calculation is performed so that the total of all the packets becomes just the full transmission path. That is, the streams are not independently transcoded, but are controlled in conjunction so that the total value of all the streams is always within the available bandwidth limit.

FIG. 8 is a schematic block diagram of a transcoder of an embodiment of the multi-program compression / encoding signal conversion apparatus according to the present invention.

As shown in FIG. 8, the present transcoder 20
0 is the input MPEG-2TS demultiplexer 210 and the output M
It includes a PEG-2TS array multiplexing unit 220, a non-reduction target TS buffer 230, a PAT / PMT generator 260, a unit time control unit 270, and a video TS processing unit 240. The video TS processing unit 240
1, video PES packet decoder 242, video PES
Packet generator 245, video TS packet generator 24
6, a video TS buffer 247 and a video transcoding unit 100. Although the video transcoding unit 100 will be described later, here, the video ES buffer 243 and the video ES transcoder 1
000. In addition, in order to realize real-time processing, each processing unit operates in parallel while ensuring synchronization.

The processing contents of each processing unit will be described below.

Input MPEG-2TS demultiplexer 210
Performs the following processing.

An input MPEG-2 transport stream is input, a system clock PCR_offset at the start of demultiplexing is obtained from a first PCR (Program Clock Reference: program clock reference value) of the input MPEG-2 transport stream, and an output MPEG signal is output. -2TS array multiplexing section 220.

The input MPEG-2TS demultiplexer 2
In step 10, the input TS packet at the unit time (n) is identified by identifying the PID of the TS packet header.
Demultiplexing is performed for each ID. Also, the type of each packet is identified by identifying the payload of the TS packet,
Processing is performed for each of the following types.

For PAT and PMT:

The above information is decoded, and each data (shown in FIG. 9) is extracted and output to the PAT / PMT generator 260. The reason why the PAT and PMT packets are not used as output TS packets as they are is to consider that when the stream transmission conditions change in the transmission path that is transcoded and output, the PAT and PMT need to be changed.

In the case of an audio stream, a system control signal other than PAT and PMT, and a null packet:

The input TS packet is directly used as the non-reduction target T
Output to the S buffer 230. Furthermore, input / output MPEG-
The non-reduction target T between two transport streams
In order to match the arrival time of the S packet to the MPEG-2 TS decoder, the non-reduction target TS within the unit time (n)
The appearance position of the packet is held and used by the output MPEG-2TS array multiplexing unit 220. I in unit time (n)
The appearance position of the non-video packet in the unit time (n)
It is represented by NonV_Run (i) using the distance between packets (the number of TS packets) from the head in the above. FIG. 10 shows a conceptual diagram of NonV_Run (i).

In the case of a video stream:

Output to the video TS packet decoder 241. However, if there is no payload even if the PID is a video stream, it is deleted.

After the above processing, if the number of input MPEG-2 TS packets satisfies the above expression (1), the non-reduction target TS
A signal indicating that the unit time (n) has elapsed is sent to the buffer 230 and the video TS buffer 247.

Also, in the unit time (0), the input MP
The PCR is decoded from the TS packet of the PCR_PID indicated in the PMT of the EG-2 transport stream, and the P
Input MPEG-2TS demultiplexer 210 before CR code
According to the total number of bytes of the MPEG-2 transport stream and the bit rate of the input MPEG-2 transport stream, the first byte is input to the input MPEG-2TS demultiplexer 210 based on the input MPEG-2 transport stream bit rate. The system clock PCR_offset at the time is calculated according to the following equation (2).

(Equation 4) First_PCR in equation (2) is the first PC encoded in the input MPEG-2 transport stream.
R, TS_cnt is the TS in which first_PCR was encoded.
Total number of input MPEG-2TS packets including packets,
PCR_point is the TS in which first_PCR was encoded.
It indicates the total number of bytes from the beginning to the byte including the last bit of the program_clock_reference_base code in the packet, and TSB_in indicates the input MPEG-2 transport stream bit rate. Also, 27,0
00,000 is the MPEG-2 reference clock 27 [MH.
z].

Next, the processing of the PAT / PMT generator 260 will be described.

When the total number TS_out of output TS packets up to the present in the unit time (n) satisfies the following equation (3), the information shown in FIG. 9 obtained from the input MPEG-2 TS demultiplexing section 210, PAT, PMT Using the information, PAT and PMT for output MPEG-2TS are newly generated and output.

(Equation 5) In Equation (3), All_TS_out (i) indicates the number of output TS packets per unit time (i), and S_out indicates the total number of PAT and PMT outputs from the start of transcoding to the present. TSB_out [bps] is the output MPEG-2
Transport stream bit rate.

Further, freq is the transmission interval of PAT and PMT.
In order to comply with the regulation that the sending interval of PCR is within 0.1 second, freq ≦ 0.1 is satisfied.

The PCR inputs the initial value to the MPEG-2TS
PCR_offset obtained by the demultiplexing unit 210 is obtained by the following equation (4).

(Equation 6) Next, the output MPEG-2TS sequence multiplexing section 220 will be described.

Output MPEG-2TS sequence multiplexing section 220
Receives a signal indicating that the non-reduction target TS buffer 230 and the video TS buffer 247 are filled, and inputs a TS packet per unit time. At this time, the input video TS in the unit time (n)
Video_TS_out (n), non-reduction target TS
Let the number of packets be NonV_TS_out (n). When the following equation (5) is satisfied, the number of PAT and PMT transmissions PAT
Here, PMT_out (n) is set to 2 here.
PATPMT_out (n) is set to 0.

(Equation 7) Equation (5) indicates whether it is necessary to output PAT and PMT in the current unit time. Where tAll
_TS_out (i) is the target output TS per unit time (i)
Shows the number of packets.

By using the above value, the number of video TS packets in the buffer per unit time (n)
TS_buff (n) is represented by the following equation (6), and the number of non-reduced TS packets in the buffer per unit time (n) No
nV_TS_buff (n) is given by the following equation (7).

Video_TS_buff (n) = Video_TS_out (n) + diffV_TS_out (n) Expression (6) NonV_TS_buff (n) = NonV_TS_out (n) + PATPMT_out (n) Expression (7)

However, diffV_TS_out (n) will be described later.

Array multiplexing of TS packets is performed by input MPEG
This is performed as follows using the information NonV_Run (i) for input / output synchronization control input from the −2TS demultiplexing unit 210.

Process 1: The target output TS packet number per current unit time target_TS_out in consideration of the difference between the past target output TS packet number and the actual output TS packet number is obtained by the following equation (8).

(Equation 8) Here, All_TS_out (i) is the actual output T per unit time (i).
This is the number of S packets.

Process 2: If the total number of output TS packets up to the present time TS_out within the unit time (n) satisfies the above equation (3), PAT and PMT are generated and output, 2 is added to TS_out, and S_out is added. to add.

Process 3: When the following expression (9) is satisfied, one video TS packet is output. If not satisfied, one non-reduction target TS packet is output, and i
Is added. Here, i indicates the number of non-reduction target TS packets output per unit time.

(Equation 9) Further, TS_out is added each time a TS packet is output.

Process 4: Return to process 2, and repeat processes 2 and 3. The termination condition is

Video_TS_buff (n) + NonV_TS_buff
If (n) is greater than target_TS_out, TS_out
= Target_TS_out, otherwise TS_out
= Video_TS_buff (n) + NonV_TS_buff (n).

After the above processing, target_TS_out <Video
In the case of _TS_buff (n) + NonV_TS_buff (n), diffV_TS_out (n + 1) is obtained by the following equation (10).

DiffV_TS_out (n + 1) = Video_TS_buff (n) + NonV_TS_buff (n) −target_TS_out Equation (10)

DiffV_TS_out (n + 1) is the excess of the total number of packets input to the output MPEG-2 array multiplexing unit by the unit time (n) with respect to the target number of output TS packets up to the unit time (n). It can be said that it is the excess of the actual output video TS packet number to the target output video TS packet number until the unit time (n). Therefore, all packets carried over as output in the next unit time as diffV_TS_out (n + 1) are video TS
Packet. Further, when the following expression (11) is satisfied, a null packet is output to the output MPEG-2TS, and a process for approaching the output bit rate is performed.

(Equation 10) STUFF_TH is a value indicating a threshold value indicating whether the stuffing process is performed when the actual output packet number is smaller than the target output TS packet number.

FIG. 11 shows an example of transition of the target time TS packet number target_TS_out including diffV_TS_out (n).

The non-reduction target TS buffer 230 receives the input M
The non-reduction target TS packets separated by the PEG-2TS demultiplexing unit 210 are held, and the non-reduction target packets are held in synchronization with the video TS buffer 247 holding the video TS transcoded by the video TS processing unit 240. T
The non-reduction target TS packet held in the S buffer is output to the output MPEG-2 TS array multiplexing unit 220.

The video TS processing section 240 receives the input MPEG
The video TS input from the −2TS demultiplexing unit 210 is decoded into a video ES, the code amount of the video ES is reduced, an output video TS is generated, and output to the output MPEG-2TS array multiplexing unit 220.

Hereinafter, each processor of the video TS processing section 240 will be described.

The video TS packet decoder 241 receives the video TS packet input from the input MPEG-2 demultiplexer 210.
And outputs the resulting video PES to the video PES packet decoder 242.

The video PES packet decoder 242 receives the video PE
S is decoded, and the obtained video ES is output to the video ES buffer 243. However, if bit segments exist before the first sequence header is detected after the start of the transcoding process, they are deleted.

Also, the input MPEG-2TS demultiplexer 2
At 10, when the unit time (n) has elapsed and a signal indicating that the buffer is empty is received from the input buffer: video ES buffer 243 of the video transcoding unit 100, the video ES in the buffer is transferred to the video ES buffer 243. Output.

Also, PTS (i), DTS (i) and PTS_DT which are synchronization information obtained at the time of video PES decoding
S_flag (i) is output to the video PES packet generator 245. Here, i indicates the picture number from the start of transcoding. Thereby, the input / output MPEG-2
The synchronization information of each picture in the transport stream can be matched.

The video PES generator 245 and the video TS packet generator 246 perform the following processing.

As soon as the video ES for one TS is generated in the video transcoding unit 100, the video transcoding unit 100 performs PES packetization and TS packetization, and stores it in the video TS buffer 247. At this time, one PES packet is composed of a video ES for one picture as shown in FIG.

The synchronization information PTS and DTS added to the PES header are the same as the video PES packet decoder 242.
DTS (i), PTS (i), PTS_DTS_
The flag (i) is used as it is in the PES header of the corresponding picture i. This ensures synchronization between the bit stream transmitted without transcoding and the bit stream whose bit rate has been reduced by transcoding.

However, since PTS and DTS are time stamps for the first access unit of the PES packet, if a PES packet of an input MPEG-2 transport stream includes a video frame of two or more pictures, the PES packet The PTS and DTS do not exist in the second and subsequent frames.

At this time, if the video PES complies with the ARIB rule that the video PES is composed of video data for one frame, the PES for the picture i having no time stamp are provided.
It is necessary to generate TS (i) and DTS (i). In that case, PTS (i) = PTS (i-1) + 90,000 / FrameRat
e, DTS (i) = DTS (i-1) + 90,000 / FrameRate.

If it is considered that the output does not need to be followed if the input does not conform to the ARIB specification, there is a method of including two frames in one PES similarly to the input MPEG-2 transport stream. .

Here, the code amount VE of the output video elementary stream up to the present within the unit time (n)
When S_out satisfies the following equation, the video transcoding unit 100 sets the time interval bf in the unit time (n).
_time of video TS packets are considered to have been packetized, and all video TS packets in the video TS buffer
The packet is output to output MPEG-2TS array multiplexing section 220.

[Equation 11] TH_out (n) shown in Expression (12) is an expected value of the total number of output bytes of the video elementary stream up to the unit time (n), and this value will be described later. Also, VES_out when Expression (12) is satisfied is changed to Video_ES_out
(N).

The video TS buffer 247 stores the video TS
The video TS packet generated by the packet generator 246 is held, and the video TS packet held in the video TS buffer is output while synchronizing with the non-reduction target TS buffer 230 holding the non-reduction target TS packet. This is output to the MPEG-2TS sequence multiplexing unit 220.

Next, the code amount control method for the multi-program transport stream will be described.

In the code amount control method of the present invention, the MPEG-2 multi-program TS bit stream for a unit time is buffered, the total target code amount of the video ES in the unit time, the picture structure of the video ES, and the constituent picture This is a method of dynamically calculating the optimal assigned code amount for each picture based on the input decoded code amount.

The features of this method are summarized below.

(1) Control is performed based on the video ES picture configuration, the input decoded code amount, and the rate distortion characteristics.

A picture structure of the video ES in a unit time,
In particular, taking into account the picture configuration for each picture type, the code amount such that the requantization error generated by transcoding is minimized based on the rate distortion characteristic function for each picture type for each program and the picture input decoding code amount. Perform control.

(2) Consider an intermediate picture at the start / end of the unit time.

For the unit time start picture, by using the same reduction rate as that of the previous unit time end picture, it is possible to suppress a large change in image quality inside the picture extending over the unit time.

(3) The rate distortion characteristic for each video ES picture is dynamically updated.

In a transcoder, unlike an encoder,
Since the coding characteristics of the video ES cannot be acquired, the dynamic characteristics fluctuate by approximately deriving the rate distortion characteristic parameter for each picture from the requantization error generated by transcoding and the reduced code amount. It is possible to flexibly deal with the image quality characteristics of the input video ES sequence.

The present system is composed of the following three steps. Step 1: Calculate the input / output code amount ratio Step 2: Calculate the reference quantization parameter Step 3: Calculate the quantization parameter based on the requantization parameter prohibition area control method

In step 1, the input decoded code amount r_s, t
target code amount R_s, t for j-th picture (picture type t {i, p, b}) in s-th program having (j)
(j) is calculated, and the ratio of the input decoded code amount to the target code amount after transcoding in the same picture (hereinafter, referred to as input / output code amount ratio ioRatio_s (j)) is calculated from the following expression (13). Aim.

IoRatio_s (j) = R_s, t (j) / r_s, t (j) Equation (13)

In deriving the target code amount R_s, t (j),
It is derived by calculating the minimum value of the quantization error generated by transcoding under the condition that the video ES total target code amount R_res is constant. Here, the total target code amount R_res of the video ES is calculated based on the target code amount of the entire TS from other target streams (such as audio streams), PMT,
It is calculated by reducing the expected code amount consumed by the PAT or the like.

Further, in step 2, the cumulative value of the input decoded code amount up to the i-th MB in the K-th picture of the s-th program, the cumulative value of the output code amount of the MB after the transcoding process, and calculated in step 1 The purpose is to calculate a reference quantization parameter using the ioRatio_s (K) value.

In step 3, the reference quantization parameter obtained in step 2 is converted into a quantization parameter that minimizes the error power when the amount of reduced codes is the same, based on the rate distortion characteristic at the time of requantization.

Hereinafter, each step will be described. (Step 1): Calculation of input / output code amount ratio

In the present invention, the input MPEG-2TS
By approximating the encoding characteristics of the video ES constituent pictures contained therein, the first stage quantization error MSE_fir_s, t
(j) and the input decoded code amount r_s, t (j) are expressed by the following equation (1).
4) shall be followed. Here, a_s, t and b_s, t are unique constants.

(Equation 12) On the other hand, the rate distortion relationship between the requantization error MSE_req_s, t (j) and the picture code amount R_s, t (j) at the time of recoding the jth picture (picture type t) in the s-th program is similarly encoded. The following equation (15) is based on the rate distortion characteristic at the time. This assumption is based on the rate-distortion characteristic at the time of requantization for the reference quantization parameter obtained in step 2 based on the rate distortion characteristic at the time of requantization so that the requantization error power is minimized when the reduced code amount is the same. It is based on making it possible to convert to a quantization parameter.

(Equation 13) From the above, the second-stage quantization error MSE_sec_s, t (j) is calculated from the re-quantization error MSE_req_s, t (j) to the first-stage quantization error M
The following equation (16) is derived as an approximation as a value obtained by subtracting SE_fir_s, t (j).

[Equation 14] From the above, the total distortion amount D_total and the video ES total target code amount R_res predicted to occur within the unit time are expressed by the following equation (17).

(Equation 15) Here, S in Expression (17) is the MPE within the unit time.
G-2 indicates the total number of video ESs present in the transport stream, and N_s indicates the total number of video ES constituent pictures in the s-th program. Further, D_s, t indicates a second-stage quantization error generated by transcoding of the j-th picture in the s-th program, and is represented by the following equation (18) from equation (16).

(Equation 16) Here, by considering the case where the last picture of the unit time is in the middle of the picture, the equation (17) is converted into the following equation (1).
Update by 9). Here, the picture type t of the last picture in unit time is set to T ({i, p, b}).

[Equation 17] In equation (19), D_s, T (R_s, T (N_s)) is an error generated by the last picture in unit time. here,
The error caused by the last picture of the unit time is predicted from the ratio of the last input picture decoding code amount r_s, T (N_s) in the current time to the input decoding code quantity μ_s, T of the picture of the same picture type T input immediately before. By doing
It is assumed that the following equation (20) holds.

(Equation 18) Next, consider the case where the remaining code of the last picture of the previous unit time exists at the start of the unit time. For the unit time start picture, the same requantization characteristic as that of the previous unit time end picture, that is, the same reduction rate ioRatio_s (n_s) is fixedly used. Here, n_s is a parameter indicating the total number of pictures N_s in the previous unit time. With this control, it is possible to suppress a large change in image quality inside a picture that extends over a unit time.

From the above, by predicting the predicted code amount of occurrence of the unit time start picture from the previous unit time ioRatio_s (n_s), it is treated as a fixed code amount independent of control, and the total target code amount R_res is calculated as follows. It is updated by equation (21).

[Equation 19] Here, in calculating the allocated code amount R_i of the I picture,
The distortion amounts D_s, p, D_s, b corresponding to the allocated code amounts R_s, p, R_s, b of the P and B pictures are the allocated code amounts R_
When converted into the amount of distortion D_s, i corresponding to s, i, the degree is estimated. Then, from R_s, p, R_s, b to R_
By performing conversion into s and i, conversion of the rate distortion function is performed. Specifically, parameters X_s, i, X_s, p, X_s, b indicating the complexity of the image used in the TM5 system are introduced, and R_s, i is determined by the ratio of X_s, i, X_s, p, X_s, b. p
(n) and R_s, b (n) are converted into R_s, i (n). The conversion formula is shown in the following formula (22).

(Equation 20) K_s, p and K_s, b in the equation (22) indicate the ratio between the average value of the I-picture decoding quantization parameter immediately before the s-th program and the average value of the immediately preceding P and B-picture decoding quantization parameters. It is expressed by equation (23).

(Equation 21) From the above, equation (19) is updated to the following equation (24).

(Equation 22) In Equation (24), N_s, i, N_s, p, N_s, b indicate the numbers of I, P, and B pictures excluding incomplete pictures at the start and end of the unit time in the video ES in the s-th program, It is expressed by the following equation (25).

N_s = N_s, i + N_s, p + N_s, b + 2 Equation (25)

On the other hand, R_res is updated by the following equation (26).

(Equation 23) Here, ψ_s and θ_s are represented by the following equations (27).

(Equation 24) Therefore, this problem is solved by the Lagrange undetermined coefficient method. λ
Is defined as the Lagrange's undetermined coefficient as in the following equation (28).

(Equation 25) In Expression (28), R_s, i (m) that satisfies ∂J / R_s, i (m) = 0 is derived. The derived result is represented by the following equation (29) and the following equation (30).

(Equation 26)

[Equation 27] From the above, R_s, i (l) (= R_s, i (m)) is calculated by the following equation (3)
It is calculated by 1).

[Equation 28] Further, R_s, p (k) and R_s, b (k) are calculated from the above-mentioned equation (22).

From the above, the input / output code amount ratio ioRatio_s (j) for each picture is calculated from equation (13). (Step 2): Reference quantization parameter calculation

At step 2, a reference quantization parameter for each MB is calculated. Specifically, in the unit time, the cumulative value of the input decoded code amount up to the i-th MB in the K-th picture of the s-th program is calculated by the equation (13) in step 1.
By feeding back the difference value between the converted virtual buffer value obtained by dividing by the ioRatio_s (j) value calculated in the above and the cumulative value of the output code amount of the MB after the transcoding process, the reference quantization scale is set. calculate.

First, i in the K-th picture of the program s
Prior to the transcoding of the MB, the occupation amount b (i) of the virtual buffer is calculated by the following equation (32).

(Equation 29) In Equation (32), B_MB_out (k, j) is the generated code amount of the j-th MB in the k-th picture per unit time, and B_MB_in (k, j)
j) indicates the decoded input code amount of the MB. Also, MB_cnt
Indicates the total number of MBs per picture. Furthermore, equation (3)
Buffer_s, t (0) in 2) indicates the initial occupancy of the virtual buffer for each picture type, and buffer_s, t (MB_cnt) in the immediately preceding picture of the same type is represented by buffer_s, t (0).
Use as

The quantization scale is calculated by the following equation (33) using the virtual buffer occupancy calculated from the equation (32).

Q_s, t (i) = 31 * buffer_s, t (i) / r Expression (33)

R in equation (33) is the same as the reaction parameter in TM5. : Rate distortion function derivation method

The value b_s, t to be derived is derived from the following equation (34) from the equation (31).

[Equation 30] (Step 3): Quantization parameter calculation based on requantization parameter prohibition area control

The reference quantization parameter Q_s, t (j) obtained in step 2 is calculated based on the requantization rate distortion characteristic,
This is a step of setting the quantization parameter to minimize the error power when the reduced code amount is the same. The details are omitted, and only the final operation expression is shown here.

If picture type t = i,

(Equation 31) For picture type t = p, b,

(Equation 32) Since it is meaningless to make the obtained Q2_s, t (j) smaller than Q1_s, t (j), if Q2_s, t (j) <Q1_s, t (j), Q2_s, Let t (j) = Q1_s, t (j). However, Expressions (35) and (36) are based on the requantization rate distortion characteristics when the requantization / quantization combination conversion is used.

The quantization using Q2_s, t (j) obtained as described above is performed.

MPEG-2 multi-program TS transcoder '1. Processor

Next, the MPEG-2 multi-program TS
FIG. 13 is a block diagram specifically illustrating the video ES transcoding unit of the transcoder 200. The TS transcoder mainly includes an MPEG-2 video ES transcoding unit and other processing units.
The MPEG-2 video ES transcoding unit is the above-described video ES transcoder 1000, and the other processing units have already been described, so description thereof will be omitted here.

The video ES transcoder 1000
LD 1110, inverse quantizer 1120, quantizer 113
0, a VLC 1140, a mode control unit 1150, and a rate control unit 1200.

[0235] Also, the video ES transcoder 1000
Decodes the video ES input by the VLD 1110 and the inverse quantizer 1120 up to the DCT coefficient region in macroblock units,
40 encodes the obtained DCT coefficient signal,
A video ES having a smaller code amount than the input video ES is generated.

In the quantization process in the quantizer 1130, the quantization process is performed by using the quantization parameter obtained by the rate control unit 1200.

As the quantization parameter, the video ES total target code amount R_ref is input to the video ES transcoder 1000, and the target code amount and the quantization parameter are calculated in the rate control unit 1200 based on R_ref. You.

Further, the rate control unit 1200
Are roughly the following five processors: a picture target code amount calculator 1210, an ioRatio calculator 1220, a quantization parameter calculator 1230, and a rate distortion parameter calculator 12
40 and a picture complexity parameter calculator 1250.

The picture target code amount calculator 1210 calculates R
_res is input, and each picture target code amount R_s, t (j) is calculated. In the calculation, processing is performed based on b_s, t input from the rate distortion parameter calculator 1240 and X_s, t input from the picture complexity parameter calculator 1250. Further, a comparison determination process is performed between the total target code amount and the total input code amount.

The ioRatio calculator 1220 receives R_res from the picture target code amount calculator 1210 and calculates ioRatio_s (j) for each picture.

The quantization parameter calculator 1230 calculates the ioRa
ioRatio_s (j) input from tio calculator 1220, VL
Based on the input decoded code amount B_MB_in (k, j) for each MB input from the D processor 1110 and the output code amount B_MB_out (k, j) for each MB input from the VLC processor 1140, reference quantization is performed. Calculate the parameters. Further, a second-stage quantization parameter Q2_s, t (i) is calculated from the first-stage quantization parameter Q1_s, t (i) and the above-described reference quantization parameter.

The rate distortion parameter calculator 1240 calculates
Restored DCT coefficient input from the first-stage inverse quantizer, VL
Q1_s, t (i) input from D processor 1110, input decoded code amount r_s, t (j), output code amount input from VLC processor 1140, and quantization parameter calculator 1230
B_s, t is calculated from Q2_s, t (i) input from.

Picture complexity parameter calculator 1250
Is the input decoded code amount r_s, t from the VLD processor 1110.
(j) and the first-stage quantization parameter Q1_s, t (i) are input, X_s, t is updated, and the picture target code amount calculator 121
Output to 0.

[0244] 2. Processing Flow: Next, flowcharts are shown in FIGS. 14 to 19, and the processing procedure will be described.

S10: Prescanning for the unit time:
The prescanning is performed on the buffered bit stream unit of the unit time portion, and the following processing is performed. s11: The input code amount r_s, t (j) in the j-th picture in the s-th program is calculated by parsing only picture_start_code. s12: Picture configuration information (N_s, N_s, i, N_s, p, N
_s, b).

S20: Calculate the total target code amount R_rec of the video ES depending on the program within the unit time.

S30: The input / output ratio ioRatio_s (j) is calculated. (Step 1) s31: Set ioRatio_s (n_s) to ioRatio_s (0). s32a: Calculate each picture target code amount R_s, t (j) based on step 1 described above. s32b: total target code amount R_tot in the s-th program
al and the total input code amount r_total are calculated by the following equation (37).

[Equation 33] s33: Comparison determination between the total target code amount and the total input code amount. s33-1: R_total_s ≦ for all programs
When r_total_s holds: The process proceeds to the following process (s34). s33-2: In cases other than s33-1: R_total_s> r
s = s *, which is _total_s, is defined as the program s *
Is subtracted from R_res.

R_res = R_res−r_total_s * Expression (38) Then, all ioRatio_s * (n) of the program s * are set as 1.
Set to 0 and return to (s31).

S34: Calculation of ioRatio_s (n): ioRatio_s (j) is calculated based on equation (13) for all pictures except the start picture in each program. At this time,
When Ratio_s (j) <1.0 holds, ioRatio_s (j) =
1.0.

S40: MB unit processing: Based on ioRatio (n) obtained in (s30), a quantization parameter for each MB is calculated and requantization processing is performed. s41: A reference quantization parameter is calculated. (Step 2) s42: A second-stage quantization parameter is calculated. (Step 3) s43: Based on the quantization parameter obtained in (s42), the MB is re-quantized. s44: The requantization square error generated by the requantization processing of (s43) is integrated.

S50: Calculation and update of X_s, t: After completion of the processing for the picture of the picture type t, the picture input code amount r_s, t (j) and the input decoding quantization parameter Q1_s, t (j) are used. , Image complexity parameter X_s, t
Is calculated by the following equation (39).

(Equation 34) s60: Substitute N_s for n_s.

S70: K_s, t is updated by equation (23).

S80: r_s, t (j), R_s, t (j) and (s
B_s, t is updated by the equation (34) from the quantization error integrated value obtained in 44). However, when the requantization process is not performed, the update is not performed.

As described above, the buffer occupancy and the quantization parameter in consideration of a plurality of programs can be obtained. By performing the quantization process using the calculated quantization parameter, the optimum code amount corresponding to the plurality of programs can be obtained. Conversion control can be performed.

[0255]

According to the present invention, an encoded signal in which a plurality of inputted programs are multiplexed is separated, and the separated first signal is obtained.
Since the code amount of each data sequence is converted for each program and the code amount conversion amount is determined in consideration of information of other programs, the code amount conversion process is performed in consideration of the overall balance of the input coded signal. And overall image quality can be improved.

[0256] In addition, by setting the target of the bit rate reduction to a video bit stream and a moving image partial stream that occupy a larger bit rate than other streams, efficient bit rate reduction can be performed.

Further, since a requantization parameter serving as a reference for code amount conversion is set for each image information, a code amount reduction rate can be set for each screen, and a code amount conversion process suitable for a screen can be performed. it can.

Further, since the conversion process of the data string is performed at a unit time interval using a predetermined unit time as a processing unit, the transmission cycle of the coded signal to be transmitted / received, for example, a drift such as a digital broadcast or a few seconds. It is possible to always perform the code amount conversion process without depending on the content, and it is possible to cope with the immediacy by adjusting the unit time.

Further, when information of one image is not input in one processing unit time, the information of the remaining images is the same as the requantization parameter used when the previous image information is subjected to code amount conversion. Is converted using the requantization parameter of, even if one image information is input over a unit time, the code amount is converted at the same information reduction rate.
It is possible to decode the image as a natural image without changing the image quality in the middle of one screen and distorting or jerking.

Also, since the requantization parameter is set by the virtual information storage amount calculated based on the code amount after the code amount conversion and the code amount before the conversion of the code amount, the converted code amount is Since the code amount reduction rate can be dynamically set in accordance with this, the optimum code amount conversion processing can always be performed.

Further, the virtual information storage amount is calculated based on an input / output code amount ratio indicating a ratio between an input decoded code amount obtained by decoding the input code amount and a target target code amount of the code amount conversion amount. Therefore, the virtual information storage amount is updated according to the decoded code amount and the target code amount, and the code amount reduction rate can be dynamically set, so that the optimum code amount conversion processing can be always performed.

Furthermore, the requantization parameter is set to the first
Since the setting is performed based on the configuration information of the image in the data sequence, the picture type, and the complexity of the image, the code amount reduction rate can be dynamically set according to the image configuration information, so that the optimal code amount conversion processing is always performed. be able to.

Further, the re-quantization parameter, in which a generated quantization error in code amount conversion processing is smaller than the reference quantization parameter, by the decoded first quantization parameter and the reference quantization parameter calculated. Is set, the quantization error generated in the code amount conversion processing can be minimized, and the optimum code amount conversion processing can be performed.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram of a rate converter showing a basic concept of the present invention.

FIG. 2 is a schematic block diagram showing features of the processing of the present invention.

FIG. 3 is a schematic block diagram of a rate converter that realizes synchronization of a video bit stream by retaining PTS and DTS of an input MPEG-2 transport stream.

FIG. 4 is a schematic diagram showing a relationship between a video TS and a non-reduction target TS in an input / output MPEG-2 transport stream.

FIG. 5 is a diagram showing a video bit stream code amount reduction mode.

FIG. 6 is a conceptual diagram illustrating a transcoding code amount control method for a VBR encoding format video bit stream.

FIG. 7 is a conceptual diagram showing a code amount control method for a plurality of programs.

FIG. 8 is a schematic block diagram showing a transcoder of an embodiment of the multi-program compression-encoding signal conversion device according to the present invention.

FIG. 9 is a diagram showing parameters of PAT and PMT.

FIG. 10 is a conceptual diagram of NonV_Run showing an appearance position of a non-video packet in a unit time (n).

FIG. 11 is a conceptual diagram showing transition of the number of unit time output TS packets.

FIG. 12 is a conceptual diagram illustrating generation of a video TS packet from a video ES.

FIG. 13 is a schematic block diagram of an MPEG-2 multi-program TS transcoder specialized for a video ES transcoding unit.

FIG. 14 is a flowchart showing a multi-program TS transcoding process.

FIG. 15 is a flowchart showing a multi-program TS transcoding process.

FIG. 16 is a flowchart showing a multi-program TS transcoding process.

FIG. 17 is a flowchart showing a multi-program TS transcoding process.

FIG. 18 is a flowchart showing a multi-program TS transcoding process.

FIG. 19 is a flowchart showing a multi-program TS transcoding process.

FIG. 20 is a schematic block diagram of a conventional transcoder.

FIG. 21 is a flowchart showing a TM5 rate control process of MPEG-2 in the conventional transcoder shown in FIG. 20;

FIG. 22 is a schematic block diagram of a conventional transcoder.

FIG. 23 is a flowchart showing a process of the conventional transcoder shown in FIG.

FIG. 24 is a schematic block diagram of a conventional transcoder.

FIG. 25 is a flowchart showing a process of the conventional transcoder shown in FIG.

[Explanation of symbols]

 Reference Signs List 50 transcoder 51 VLD (variable length decoding means) 53 inverse quantizer (inverse quantization means) 55 quantizer (quantization means) 57 VLC (variable length encoding means) 59 rate control unit 60 transcoder 61 delay circuit 63 Bit rate ratio calculation unit 65 Input code amount integration unit 67 Difference code amount calculation unit 69 Target output code amount update unit 71 Quantization scale code calculation unit 80 Transcoder 81 VLD 83 Target output code amount update unit 85 Quantization scale code calculation Unit 100 Video transcoding unit 200 Transcoder (encoded signal conversion device) 210 Input MPEG-2TS demultiplexing unit 220 Output MPEG-2TS array multiplexing unit 230 Non-reduction target TS buffer 240 Video TS processing unit 241 Video TS packet decoder 242 Video PES packet recovery 243 Video ES buffer 245 Video PES packet generator 246 Video TS packet generator 247 Video TS buffer 260 PAT, PMT generator 270 Unit time control unit 600 Rate converter 610 MPEG-2TS demultiplexer 620 MPEG-2TS multiplexer 640 MPEG-2 video transcoder 650 System controller 700 Rate converter 710 MPEG-2TS demultiplexer 720 MPEG-2TS multiplexer 741 Video TS decoder 742 Video PES decoder 744 Video ES transcoder 745 Video PES packet generator 746 Video TS packet generator 1000 Video ES transcoder 1110 VLD 1120 Inverse quantizer 1130 Quantizer 1140 VLC 1150 Mode control unit 1 00 rate controller 1210 picture target code amount calculator 1220 IoRatio calculator 1230 quantization parameter calculator 1240 rate distortion parameter calculator 1250 picture complexity parameter calculator

 ──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Maki Sugiura 2-4-12 Okubo, Shinjuku-ku, Tokyo Inside Media Glue Co., Ltd. (72) Inventor Hiroyuki Kasai 1-3-1-10 Nishiwaseda, Shinjuku-ku, Tokyo Inada University International Information and Communication Research Center (72) Inventor Hideyoshi Tominaga 1-3-10 Nishiwaseda, Shinjuku-ku, Tokyo Waseda University International Information and Communication Research Center F-term (reference) 5C059 KK34 KK41 MA00 MC11 ME01 RB02 RB10 SS02 TA60 TC00 TC18 5J064 AA01 AA05 BA09 BA16 BC01 BC16 BD02 5K028 AA12 AA14 EE03 EE07 KK01 KK03 KK32 MM05 5K034 AA05 HH12 HH21 HH61 JJ23 KK28 MM02 MM14 MM31

Claims (30)

[Claims] An input step of inputting, via a first transmission line having a first transfer rate, a first encoded signal obtained by multiplexing and compressing and encoding a plurality of programs; A signal demultiplexing step of separating the first encoded signal into a first data sequence, a second data sequence, and a third data sequence; and a plurality of first data sequences separated in the signal demultiplexing step, A data string converting step of generating a plurality of converted first data strings having a smaller total code amount than a total code amount of the first data string; and converting the third data string separated in the signal demultiplexing step into a first encoded signal. A third data string correction step of correcting the data according to the characteristic change after the conversion to generate a corrected third data string; a conversion first data string generated in the data string conversion step; Minute A signal array multiplexing step of multiplexing the separated second data string and the corrected third data string corrected in the third data string correction step to generate a second encoded signal; and the first transfer Outputting the second encoded signal via a second transmission path having a second transfer rate lower than the speed, wherein the data string conversion step converts the first data string into the first data string. When performing code amount conversion on a data sequence, multi-program compression is performed, wherein the code amount conversion of the first data sequence is performed for each program, and information of another program is also acquired to determine the code amount conversion amount. Coded signal conversion method. 2. The multi-program compression coded signal conversion method according to claim 1, wherein said inputting step comprises the step of:
A G-2 transport stream, wherein the signal demultiplexing step separates a transport stream packet including a compression-encoded video signal as the first data stream; and the output step includes the second encoding. MPE as signal
A multi-program compression-encoded signal conversion method, comprising outputting a G-2 transport stream. 3. The multi-program compression coded signal conversion method according to claim 1, wherein said inputting step inputs a multiplexed audio / video compression coded stream as said first coded signal, and said signal demultiplexing step. Separates a moving image partial stream as the first data stream, and the output step outputs a multiplexed audio / video compression-encoded stream as the second encoded signal. Signal conversion method. 4. The multi-program compression / encoding signal conversion method according to claim 2, wherein said data string conversion step comprises: for each image information of said first data string, a reference of a conversion amount of said code amount conversion. A multi-program compression-encoded signal conversion method, comprising: setting a re-quantization parameter, and converting a code amount of the image information according to the re-quantization parameter. 5. The multi-program compression coded signal conversion method according to claim 4, wherein all data string conversion processes for generating the second coded signal from the first coded signal are performed in a predetermined unit time. Performing the interval as a processing unit, wherein the data string conversion step inputs the first data string in the unit time interval, and the first data string based on information of the first data string in the unit time. A multi-program compression-encoded signal conversion method, each of which determines a code amount conversion amount. 6. The multi-program compression coded signal conversion method according to claim 5, wherein the data string conversion step is performed when one image information of the first data string is input over the unit time. In the unit time when the remaining image information part of the image information is input, the remaining image information part is subjected to code amount conversion according to a requantization parameter used when converting the converted image information. Multi-program compression coded signal conversion method. 7. The multi-program compression / encoding signal conversion method according to claim 4, wherein the data string conversion step includes a step of inputting image information to be subjected to code amount conversion within a unit time. The virtual information storage amount is calculated based on the code amount before the image information and the converted code amount of the image information before the image information, and the requantization parameter is set based on the virtual information storage amount. A multi-program compression-encoding signal conversion method characterized by the above-mentioned. 8. The multi-program compression / encoding signal conversion method according to claim 7, wherein the data string conversion step includes an input decoding code amount obtained by decoding the input code amount of the image information, and a code amount conversion of the image information. A multi-program compression-encoding signal conversion method, wherein the virtual information storage amount is calculated based on an input / output code amount ratio that is a ratio of a target amount of code to a target code amount. 9. The multi-program compression / encoding signal conversion method according to claim 4, wherein said data string conversion step includes a step of converting the image information of said first data string into a code amount. A multiprogram compression-encoded signal conversion method, wherein the reference requantization parameter is set based on configuration information of the image. 10. The multi-program compression-encoded signal conversion method according to claim 4, wherein said data string conversion step is obtained by decoding image information of said first data string. A quantization error generated in a code amount conversion process using the reference quantization parameter from a first quantization parameter when encoding image information and a reference quantization parameter that is a reference of code amount conversion calculated by calculation. Wherein the re-quantization parameter is set such that a quantization error that occurs is smaller than that of the re-quantization parameter. 11. A first coded signal having a plurality of programs multiplexed and compressed and coded is converted to a first coded signal having a first transfer rate.
Input means for inputting via a transmission path, signal demultiplexing means for separating the first coded signal input by the input means into a first data sequence, a second data sequence, and a third data sequence; A data string converting means for generating, from a plurality of first data strings separated by the demultiplexing means, a plurality of converted first data strings having a total code amount smaller than a total code amount of the plurality of first data strings; A third data string modifying unit that modifies the third data string separated by the demultiplexing unit according to a characteristic change after the conversion of the first encoded signal to generate a modified third data string; Multiplexing the converted first data string generated by the means, the second data string separated by the signal demultiplexing means, and the corrected third data string corrected by the third data string correcting means; Signal array multiplexing means for generating two encoded signals And output means for outputting the second encoded signal via a second transmission path having a second transfer rate lower than the first transfer rate, wherein the data stream conversion means comprises: When performing the code amount conversion of the sequence to the first data sequence, the code amount conversion of the first data sequence is performed for each program, and the information of another program is also acquired to determine the code amount conversion amount. A multi-program compression-encoded signal conversion device characterized by the following. 12. The multi-program compression-encoding signal conversion apparatus according to claim 11, wherein said input means comprises an MPEG-coded signal as said first encoded signal.
2 transport stream, the signal demultiplexing unit separates a transport stream packet including a compression-encoded video signal as the first data stream, and the output unit outputs the second encoded stream as the second encoded signal. MPEG-
A multi-program compression-encoding signal conversion device for outputting two transport streams. 13. The multi-program compression coded signal conversion apparatus according to claim 11, wherein said input means inputs a multiplexed audio / video compression coded stream as said first coded signal, and said signal demultiplexing means. Separates a moving image partial stream as the first data stream, and the output means outputs a multiplexed audio / video compression-encoded stream as the second encoded signal. Signal converter. 14. The multi-program compression-encoding signal conversion device according to claim 12, wherein said data sequence conversion means sets a reference for a conversion amount of said code amount conversion for each image information of said first data sequence. A multi-program compression-encoded signal conversion apparatus, wherein a re-quantization parameter is set and a code amount of the image information is converted according to the re-quantization parameter. 15. The multi-program compression coded signal conversion device according to claim 14, wherein all data string conversion processes for generating the second coded signal from the first coded signal are performed in a predetermined unit time. The interval is used as a processing unit, and the data string conversion means inputs the first data string in the unit time interval, and the first data string based on information of the first data string in the unit time. A multi-program compression-encoded signal conversion device, each of which determines a code amount conversion amount. 16. The multi-program compression-encoding signal conversion apparatus according to claim 15, wherein said data string conversion means has been converted when one image information of said first data string is input over said unit time. In the unit time when the remaining image information part of the image information is input, the remaining image information part is subjected to code amount conversion according to a requantization parameter used when converting the converted image information. Multi-program compression-encoded signal conversion device. 17. The multi-program compression-encoding signal conversion apparatus according to claim 14, wherein said data string conversion means inputs image information to be converted in a code amount within a unit time. Code amount before the image information
Calculating a virtual information storage amount based on the converted code amount of the image information before the image information, and setting the requantization parameter based on the virtual information storage amount. Signal conversion device. 18. The multi-program compression-encoding signal conversion apparatus according to claim 17, wherein said data string conversion means decodes an input code amount of said image information and converts a code amount of said image information. A multi-program compression-encoded signal conversion apparatus, wherein the virtual information storage amount is calculated based on an input / output code amount ratio that is a ratio of a target amount of code to a target amount of code. 19. The multi-program compression-encoded signal conversion apparatus according to claim 14, wherein said data sequence conversion means converts the image information of said first data sequence into a code amount. A multi-program compression-encoded signal conversion apparatus, wherein the reference re-quantization parameter is set based on configuration information of the image. 20. The multi-program compression-encoded signal conversion device according to claim 14, wherein said data string conversion means decodes image information of said first data string. First when encoding image information
A quantization error generated from the quantization parameter and a reference quantization parameter serving as a reference of the code amount conversion calculated by calculation is smaller than a quantization error generated in the code amount conversion process using the reference quantization parameter. A multi-program compression-encoding signal conversion device, wherein the re-quantization parameter is set. 21. A first coded signal having a plurality of programs multiplexed and compressed and coded is converted to a first coded signal having a first transfer rate.
An input step of inputting via a transmission path, a signal demultiplexing step of separating the first encoded signal input in the input step into a first data sequence, a second data sequence, and a third data sequence; A data string conversion step of generating a plurality of converted first data strings having a smaller total code amount than a total code amount of the plurality of first data strings from the plurality of first data strings separated in the demultiplexing step; A third data string modification step of modifying the third data string separated in the demultiplexing step according to a characteristic change after the conversion of the first encoded signal to generate a modified third data string; Multiplexing the converted first data string generated in the step, the second data string separated in the signal demultiplexing step, and the corrected third data string corrected in the third data string correcting step; A signal array multiplexing step of generating two encoded signals; and an output step of outputting the second encoded signals via a second transmission path having a second transfer rate lower than the first transfer rate. When the data string conversion step performs the code amount conversion of the first data string to the converted first data string, the data string conversion step performs the code amount conversion of the first data string for each program, and also obtains information of another program. A medium for recording a multi-program compression-encoded signal conversion program, wherein the code amount conversion amount is determined. 22. A medium on which a multi-program compression-encoded signal conversion program according to claim 21 is recorded, wherein said inputting step comprises the step of:
A G-2 transport stream, wherein the signal demultiplexing step separates a transport stream packet including a compression-encoded video signal as the first data stream; and the output step includes the second encoding. MPE as signal
A medium on which a multi-program compression-encoded signal conversion program for outputting a G-2 transport stream is recorded. 23. A medium recording a multi-program compression-coded signal conversion program according to claim 21, wherein said input step inputs a multiplexed audio / video compression-coded stream as said first coded signal, The signal demultiplexing step separates a moving image partial stream as the first data stream, and the output step outputs a multiplexed audio / video compression encoded stream as the second encoded signal. A medium on which a program compression coded signal conversion program is recorded. 24. A medium on which a multi-program compression-encoded signal conversion program according to claim 22 or 23 is recorded, wherein said data string conversion step comprises: converting said code amount conversion for each image information of said first data string. A medium storing a multi-program compression / encoding signal conversion program, wherein a requantization parameter serving as a reference of an amount is set, and the image information is code-quantized according to the requantization parameter. 25. A medium on which a multi-program compression-coded signal conversion program according to claim 24 is recorded, wherein all data string conversion processing for generating the second coded signal from the first coded signal is performed. While performing a predetermined unit time interval as a processing unit, the data string conversion step inputs the first data string in the unit time interval, and based on information of the first data string in the unit time, A medium storing a multi-program compression / encoding signal conversion program, wherein a code amount conversion amount of a first data string is determined. 26. A medium on which a multi-program compression / encoding signal conversion program according to claim 25 is recorded, wherein said data string conversion step inputs one image information of said first data string over said unit time. When the remaining image information portion of the converted image information is input in a unit time, the remaining image information portion is subjected to code amount conversion according to the requantization parameter used when converting the converted image information. A medium on which a multi-program compression-encoded signal conversion program is recorded. 27. A medium on which a multi-program compression-encoded signal conversion program according to claim 24 is recorded, wherein said data string conversion step includes a step of converting the image information to be subjected to code amount conversion into a unit time. The virtual information storage amount is calculated based on the code amount before the image information and the code amount after the conversion of the image information before the image information, and the requantization is performed based on the virtual information storage amount. A medium for recording a multi-program compression-encoded signal conversion program characterized by setting parameters. 28. A medium having recorded thereon a multi-program compression-encoding signal conversion program according to claim 27, wherein said data string conversion step comprises: an input decoding code amount obtained by decoding an input code amount of said image information; A multi-program compression-encoded signal conversion program characterized in that the virtual information storage amount is calculated based on an input / output code amount ratio which is a ratio of a target code amount targeted for the code amount conversion amount. Medium. 29. A medium on which the multi-program compression-encoded signal conversion program according to claim 24 is recorded, wherein said data string conversion step includes converting the image information of said first data string into a code amount. A medium storing a multi-program compression-encoded signal conversion program, wherein the re-quantization parameter serving as a reference for conversion is set based on configuration information of the image. 30. A medium recording the multi-program compression-encoded signal conversion program according to claim 24, wherein said data string conversion step decodes image information of said first data string. Generated in a code amount conversion process using the reference quantization parameter, from a first quantization parameter when the image information obtained by encoding is obtained, and a reference quantization parameter that is a reference of code amount conversion calculated by calculation. A medium storing a multi-program compression-encoded signal conversion program, wherein the re-quantization parameter is set such that a quantization error generated is smaller than a quantization error to be generated.