JP2003224847A - Image encoder and method thereof, and image decoder and method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image encoder and a method thereof capable of encoding many image signals with a motion such as HDTV signals with high image quality without a time delay and to provide an image decoder and a method thereof. <P>SOLUTION: A processor 130<SB>-1</SB>in a division image encoder 120<SB>-1</SB>encodes two-dimensional sub image signals resulting from applying 1/N division to a two-dimensional image signal by an image division device 110 and calculates a feature parameter denoting the complicatedness of the image on the basis of the encoded result. A share bit rate arithmetic section 171 calculates a proper bit rate to encode an image signal of a succeeding frame on the basis of a temporal change in the feature parameter denoting the complicatedness of the image and shares the bit rate to a plurality of the processors 13<SB>-1</SB>. A plurality of the processors 130<SB>-1</SB>encodes sub image signals at a designated bit rate. A video stream integral device 160 integrates the encoded sub image signals to generate and output the video stream. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION The present invention relates to, for example, an HDTV.
The present invention relates to an image encoding apparatus and method for compressing and encoding a moving image signal such as a (High Definition TeleVision) signal, an image decoding apparatus and method for decoding the encoded signal.

[0002]

2. Description of the Related Art MPEG2 is used as a moving image coding system.
The standard (ISO / IEC13818) is widespread. In MPEG2, the concept of a profile that mainly defines the classification of functions (difference in syntax) and a level that defines the difference in the amount of processing such as image size is introduced, and the coding performance that can be supported is classified into classes. There is.
For example, for an image of ITU-R601 with 720 x 480 pixels and 60 fields / sec, MP @ ML (Main Pr
ofile at Main Level) is used, 1920 × 1
MP @ HL (Main Profile at High Level) is used for an HDTV image of 080 pixels and 60 fields / second.

An apparatus for encoding or decoding a moving image signal having a screen configuration with a large number of pixels such as an HDTV signal, that is, an apparatus for performing encoding and decoding of MP @ HL, for example, is very high speed processing. Is difficult to realize, and it is a very expensive device considering the realization. On the other hand, a device that executes MP @ ML encoding and decoding may be smaller and have a lower operation speed than a device of MP @ HL, and it is widely used because it has already been integrated into an LSI. Therefore, it can be used very inexpensively.

In the moving picture coding / decoding apparatus described in Japanese Patent Application Laid-Open No. 10-234043, a moving picture signal having a screen configuration with a large number of pixels, such as an HDTV signal, is divided into a plurality of screens. A method has been proposed in which signals of respective screens are encoded or decoded by a device of MP @ ML and the processing results are integrated to perform encoding or decoding of MP @ HL.

[0005]

However, in the moving picture coding apparatus described in Japanese Patent Application Laid-Open No. 10-234043, coding is performed so that the amount of codes generated in each divided area of a moving picture is uniform. Because it controls
If the complexity of the image of each divided screen varies, the code amount assigned to the screen of the complex image is not sufficient, and the image quality deteriorates.

In order to solve such a problem, by setting values such as a target bit, a maximum bit and a minimum bit for each frame for each of a plurality of MP @ ML encoding devices, it is possible to obtain an appropriate value. It is necessary to allocate the code amount. However, such control is complicated in processing, and is very small from the time when the coding of the previous picture is completed and the information such as the code amount is obtained to the time when the coding of the next incoming picture is started. This is a very difficult process that requires high-speed processing because it must be executed within a short time. Further, in such a configuration, the maximum code amount generated for each split screen must be considered as the maximum code amount for the entire screen. It is necessary to provide a buffer having a capacity capable of buffering the code of the entire screen in the preceding stage of the MP @ ML decoding device of No. 3, which causes a problem that the capacity of the buffer increases.

When one moving image signal is encoded using a plurality of encoders, for example, Japanese Patent Laid-Open No. 11-25255.
As disclosed in Japanese Patent Publication No. 0, VBV (Video B
A single buffer is used. If a single VBV buffer is to be used, the control means for controlling the entire coding device is virtually a single VBV buffer.
Having an image of a V buffer, and obtaining information on the amount of data stored in that buffer from each of a plurality of encoders,
Based on this, the occupancy of the buffer is detected, and based on this, rate control is performed for each encoder. Also in this case, the transfer and calculation of these pieces of information must be performed in a very short time after the coding of one picture is completed and before the coding of the next picture is started. There is a problem that ability is required.

A method called statistical multiplexing is known as a method for multiplexing bit streams coded at a variable bit rate by a plurality of encoders and collectively transmitting them as one constant bit rate bit stream.
Particularly, a method for appropriately controlling the amount of data within the capacity of a transmission line or an encoder or a decoder is disclosed in International Publication W098 / 32252. However, when dividing a moving image signal of HDTV, encoding each as a variable bit rate by an encoder for SDTV, and integrating a plurality of output bitstreams into one bitstream of a constant bitrate, this is It is different from multiplexing, and the same method as statistical multiplexing cannot be applied.

Japanese Unexamined Patent Publication No. 2000-224589 discloses a method of appropriately allocating a bit rate to a plurality of encoders (encoders) in a technique of dividing an image and encoding it. However, this method is intended to be applied to a storage recording medium and employs a two-pass encoding method. Therefore, the encoding time becomes long and it is not suitable for the purpose of encoding in real time. In addition, the bit rate is not constant, but a variable rate is assumed, and the processing is complicated.

As a method for overcoming the above-mentioned problems, the inventor of the present application has disclosed in the specification and drawings of Japanese Patent Application No. 2000-205115 (filed on Jul. 6, 2001) the bit by the feature amount of a picture already encoded. We have proposed an invention that allocates rates. However, according to the present invention, a time delay occurs in the processing for an image whose image feature amount changes with time. Therefore, the present invention has found that the bit rate is not necessarily allocated to an appropriate bit rate.

An object of the present invention is to appropriately allocate a bit rate to a moving image signal such as an HDTV which has a large number of pixels and whose image feature amount changes with time, and encodes it with high image quality. An object of the present invention is to provide an image encoding device and an image encoding method that can do the above. Another object of the present invention is to provide an image decoding device and an image decoding method capable of decoding a moving image signal thus encoded with high image quality.

[0012]

According to a first aspect of the present invention, image signal dividing means for dividing a two-dimensional image signal into N two-dimensional sub-image signals, and the image signal dividing means are divided into N pieces. N encoding means for encoding each of the plurality of generated two-dimensional sub-image signals at a specified bit rate, and a characteristic parameter indicating the complexity of the image of the sub-image signal is calculated from the encoding result. And a bit rate calculation means for allocating a bit rate to be used in each of the N encoding means from the calculated N feature parameters, and an encoding by the N encoding means. Coded image signal integration means for integrating the N encoded sub-picture signals, each of the N encoding means according to the bit rate calculated by the allocation bit rate calculation means. The divided two-dimensional sub-image signals are encoded, and the characteristic parameter calculation means calculates a characteristic parameter indicating the complexity of each image of the two-dimensional sub-image signals from the encoded result. A device is provided.

Preferably, the encoding means is characterized by a characteristic parameter Xi, n (i = 1 ,.
, N, and n indicates the n-th frame) is calculated based on the following equation.

[0014]

Xi, n = Si, n × Qi, n where Si, n is the generated code bit amount of n frames of the sub-image signal i, and Qi, n is the average quantum of n frames of the sub-image signal i. It is a scale.

[0015]

Xi, n = Σ (Si, j × Qi, k) where Si, j is the generated code bit amount for each j frame of the sub-image signal i (j = nl, ..., n) -1, n) Qi, k is an average quantization scale for each k frame of the sub-image signal i (k = nm, ..., N-1, n).

[0016]

Xi, n = Σ (Si, j × Qi, k) / (N−1) where Si, j is the generated code bit amount of the j-frame image of the sub-image signal i (j = nl, ..., n-1, n) Qi, k is an average quantization scale of k frames of the sub-image signal i (k = nm, ..., n-1, n).

Preferably, the allocation bit rate calculation means calculates a predicted value of the characteristic parameter concerning the complexity of the image of the sub-image signal to be encoded next from the temporal change of the characteristic parameter, and the calculated predicted characteristic. The allocated bit rate is calculated from the parameters.

More preferably, the allocation bit rate calculating means obtains the predicted value (X'i, n + 2) of the characteristic parameter of the (n + 2) th frame by the following formula.

[0019]

X'i, n + 2 = (Xi, n) +2 * [(Xi, n-Xi, n-1)] = 3 * (Xi, n-2) -2 * (Xi, n- 1) where Xi, n is a characteristic parameter of the sub-image signal i of the nth frame, and Xi, n-1 is a characteristic parameter of the sub-image signal i of the (n-1) th frame.

Further preferably, the allocation bit rate calculating means obtains the predicted value (X'i, n + 1) of the characteristic parameter of the (n + 1) th frame by the following formula.

[0021]

X'i, n + 1 = Xi, n + (Xi, n-Xi, n-1) = 2 * (Xi, n)-(Xi, n-1) where Xi, n is n Xi, n-1 is a feature parameter of the sub-image signal i of the (n-1) th frame.

More preferably, the allocation bit rate calculating means calculates the allocation of the bit rate R from the prediction feature parameter (X'i, n + 1) regarding the complexity of the image based on the following equation.

[0023]

[Equation 18] Ri, n + 1 = R × (X'i, n + 1) / Σ (X'i, n + 1)

Preferably, each of the N coding means has a buffer for storing a video stream generated by the coding. Also preferably, the N
Each of the encoders has a VBV (Video Buffering Ver.
ifier) buffer, each of the N encoders having
The encoding is performed based on the occupied amount of the VBV buffer provided.

According to a second aspect of the present invention, there is provided an image coding method for implementing the above image coding apparatus. That is, the image coding method uses a two-dimensional image signal N
An image signal dividing step of dividing into two two-dimensional sub-image signals, an encoding step of encoding the N divided two-dimensional sub-image signals at a designated bit rate, and the encoding step From the result, a characteristic parameter indicating the complexity of the image of each two-dimensional sub-image signal is calculated, and a bit rate in the encoding step is calculated from the calculated characteristic parameter, and the calculation is performed. The bit rate calculation step for reflecting the bit rate in the processing of the encoding step, and the encoded image signal integration step for integrating the two-dimensional sub-image signals encoded in the encoding step. ,, in the encoding step, encoding the next plurality of divided two-dimensional sub-image signals according to the bit rate calculated in the allocation bit rate calculation step,
In the characteristic parameter calculation step, a characteristic parameter indicating the complexity of the image of the two-dimensional sub-image signal is calculated from the encoded result.

According to a third aspect of the present invention, there is provided an image decoding device comprising a decoding means for decoding an image signal obtained by the image coding device or the image coding method according to the bit rate. To be done.

According to a fourth aspect of the present invention, there is provided an image decoding method having a step of decoding an image signal obtained by the image coding apparatus or the image coding method according to the bit rate. .

The image signal of the image decoding apparatus and method divides the two-dimensional image signal into N two-dimensional sub-image signals, and divides the N divided two-dimensional sub-image signals. Encoding is performed at a specified bit rate, a characteristic parameter indicating the complexity of the image of each two-dimensional sub-image signal is calculated from the encoded result, and used in the encoding process from the calculated characteristic parameter. It is an image signal obtained by calculating the bit rate and integrating the two-dimensional sub-image signals encoded by the encoding process.

[0029]

BEST MODE FOR CARRYING OUT THE INVENTION First Embodiment Referring to FIGS. 1 to 7, an image coding apparatus and an image coding method, an image decoding apparatus and an image decoding method according to the present invention,
In addition, a first embodiment of the image transmission device will be described. The image encoding device will be described. FIG. 1 is a block diagram showing the configuration of the image encoding device 101 according to the first embodiment. The image encoding device 101 is the image division device 1.
10, first to fourth divided image coding devices 120 _{-1 to} 12
_0-4 , video stream integration device 160 and control device 170. The image dividing device 110 corresponds to the image signal dividing means for dividing the two-dimensional image signal of the present invention into N two-dimensional sub-image signals. The first to fourth divided image coding devices 120 _{-1 to} 120 _-4 correspond to the N coding means of the present invention. Video stream integration device 1
Reference numeral 60 corresponds to the encoded image signal integrating means of the present invention. The controller 170 includes the bit rate allocation means of the present invention.

The preferred image coding apparatus 101 according to the present embodiment shows a preferred embodiment for decoding a large amount of image signals such as HDTV signals. In such a case, When one encoding means performs one frame of HDTV signal, the encoding process takes a very long time,
Since it is difficult to realize the encoding process in real time, the sub-image signal obtained by dividing the image signal into N pieces is encoded. Hereinafter, in the present embodiment, the number N indicating a plurality of divisions is 4.

As shown in FIG. 2, the image dividing device 110 converts each frame image of a continuously input two-dimensional moving image signal into four areas A, B, C and D.
The image data is divided into four sub-image signals, and the four divided sub-image signals are divided into first to fourth divided image coding devices 120 _{-1 to} 120 _-4.
Output to. Details of the processing of the image division device 110 will be described.

The two-dimensional moving image signal input to the image encoding device 101 illustrated in FIG. 2 is, for example, an HDTV video signal, the luminance signal is 1920 pixels horizontal × 1080 lines vertical, and the color difference signal is 960 pixels horizontal × It is a 4: 2: 2 format interlaced signal with vertical 1080 lines. The frame rate is 30 frames / second. Image dividing device 110
Converts the luminance signal into 1440 pixels in the horizontal direction and the color difference signal into 720 pixels in the horizontal direction by filtering. The vertical direction is determined by the MPEG2 standard to be a multiple of 32 lines in an interlaced image, so that 8 lines of dummy data are added below the image for both the luminance signal and the color difference signal to form 1088 lines. . The image division device 110 divides the image whose size has been converted in this way into four areas A, B, C, and D as shown in FIG. 2, and the sub-image signals of the respective divided areas are divided into first to fourth areas. To the divided image coding devices 120 _{-1 to} 120 _-4 . For example, the luminance signal is divided into four sub-image signals of horizontal 720 pixels × vertical 544 lines.

The first to fourth divided image coding devices 120 _-1
Each of 120 to _-4 is a bit rate designated by the control device 170, encodes a sub-image signal input from the image division device 110, and a characteristic parameter indicating the image complexity of the sub-image signal from the encoded result. To calculate. Also, the first
~ The fourth divided image coding devices 120 _{-1 to} 120 _-4 buffer the coded results and output them to the video stream integration device 160 at a predetermined rate. Each divided image coding device 120 _-i (i = 1 to 4) has a processor 130 _-i and a buffer 150 _-i .

The processor 130 _-i is, for example, an MPE.
It is a G2-MP @ ML encoder that encodes an input sub-image signal by the MPEG2 method. The video stream generated by the encoding is output to the buffer 150 _-i substantially instantaneously in picture units. Since the NTSC system may support up to 480 vertical lines, the processor 130 _{-i according to the} present embodiment has a vertical direction of 57 lines.
Encode in PAL mode that supports up to 6 lines. Further, in the PAL mode, since the frame rate is 25 frames / second, the clock frequency is increased and encoding is performed at 30 frames / second.

The processor 130- _i MPs the sub-image signal.
A characteristic parameter indicating the image complexity of the sub-image signal encoded by the EG2 method is calculated. As the characteristic parameter indicating the complexity of the image, any parameter may be used as long as it satisfies the requirement, but in the present embodiment, as a preferred first example, the complexity of the image of the sub-image signal i is Characteristic parameter Xi, n (i is the number of the sub-image signal, and n
Is the frame number of the image signal. ), The product of the amount of generated bits Si, n obtained by encoding for each picture and the average quantization scale Qi, n defined by the following equation is used.

[0036]

Xi, n = Si, n × Qi, n (1) where Si, n is the generated code bit amount for each frame of the sub-image signal i, and Qi, n is the sub-image It is the average quantization scale for each frame of signal i.

As the characteristic parameter Xi, n indicating the complexity of the image of the sub-image signal i, in addition to the above-mentioned examples, for example, with the intention of canceling the difference due to the picture type, It is also possible to use the total of the past three pictures, or the average as shown in the following Expression 3.

[0038]

[Equation 20]     Xi, n = Si, n x Qi, n + Si, n-1 xQi, n-1 + Si, n-2 xQi, n-2                                                       ... (2)

[0039]

[Equation 21]     Xi, n = [Si, n x Qi, n + Si, n-1 x Qi, n-1 + Si, n-2 x Qi, n-2]               / 3                                                       ... (3)

The buffer 150 _-i is the processor 130 _-i.
The coded video stream input from the device is temporarily stored, and the allocation bit rate calculation unit 17 of the control device 170 is stored.
1 is output to the video stream integration device 160 at the bit rate instructed. Allocation bit rate calculation unit 1
The process of 71 will be described later. From the first to the fourth divided image coding devices 120 _{-1 to} 120 _-4 , respectively, MPEG
A 2-MP @ ML video stream is output. The format of the moving image signal input to the image encoding device 101 is 4:
2: 2, but in processor 130- _i 4: 2:
Since the video stream is converted into the 0 format, each of the video streams output from the first to fourth divided image coding apparatuses 120 _{-1 to} 120 _-4 is a video stream of a moving image signal in the 4: 2: 0 format. is there.

The video stream integrating device 160 has a first
~ Fourth divided image encoding device 120 _-1~ 120_-FourOut of
Four MPEG2-MP @ ML video streams supported
Video stream of MPEG2-MP @ HL
Generate a game. That is, the video stream integration device
160 is four MPEG2-MP @ ML video streams.
Break the reme into slices and put them in one video
One MPEG2 by reconstructing into a stream
-Generate MP @ HL video stream. At this time,
Sequences higher than the slice unit, GOP (Group Of
 Picture), picture-level header data, extended data
Since only one user data and user data are required,
The system integration device 160 uses the first divided image encoding device 12
0_-1Video stream output from (split image encoding device A)
Second to fourth divided image coding apparatus using only stream
120_-2~ 120_-FourFrom (divided image coding devices B to D)
The output video stream is discarded. More specifically
To be more specific, the video stream integration device 160 includes the first to the first
Fourth divided image coding device 120_-1~ 120_-FourIn
Processor 130_-1~ 130_-FourOpen the encoding process in
After the start, four divided image coding devices 120_-1~ 120_-FourOr
In each of the four video streams output from
First slice until the first slice start code appears
Divided image coding device 120_-1(Divided image coding device A)
Output only the video stream output from
Fourth divided image coding device 120_-2~ 120_-Four(Split image
Video stream output from image encoding devices B to D)
Discard.

At the same time, one MPEG2-MP @ H
Parameters that need to be rewritten as an L video stream, such as horizontal pixel size, vertical pixel size, aspect ratio information, bit rate, VBV buffer size, and VBV delay, follow instructions from the control unit 172 of the control device 170, or What is possible is rewritten to the value obtained by calculation from each value of the video stream output from each divided image encoding device 120 _-i (i = 1 to 4). It should be noted that if the F code representing the search range of the motion vector is rewritten to the same value in picture units in the processors 130 _-1 of the first to fourth divided image coding devices 120 _{-1 to} 120 _-4 , the F code Since the conversion and rewriting of the vector value are required and the processing is complicated, the control unit 172 causes the processor 130 _-1 of each divided image encoding device 120 _-i (i = 1 to 4) to encode each picture. It is desirable to set the same value in advance.

The video stream after the slice start code is decomposed in units of slices, and as shown in FIG. 3, A1, B1, A2, B2, ..., A34, B34, C1,
D1, C2, D2, ..., C34, D34 are rearranged in this order and output. At this time, the value of the slice header indicating the vertical position information of the slice is rewritten to a correct value as necessary. Also, the macro block address increment representing the horizontal position information in the first macro block of the slice is rewritten to a correct value as necessary.

After outputting the video stream of the slice D34, the video stream integrating device 160 outputs the sequence end code to 1 when the sequence end code is output from the divided image coding devices 120 _-i (i = 1 to 4).
One is output, and synthesis is completed or the next sequence start code is input. If the sequence end code is not output, the next picture start code or group start code (start code of GOP) is output, and the video stream integration device 160 repeats similar processing.

The control device 170 has an allocation bit rate calculation unit 171 and a control unit 172. Control unit 17
2 controls the operations of the allocation bit rate calculation unit 171, the image division device 110, the divided image encoding devices 120 _{-1 to} 120 _-4 , and the video stream integration device 160.

The allocation bit rate operation unit 171 is
The above-mentioned feature parameter Xi, which indicates the complexity of the image, output from one of the divided image coding devices 120 _{-1 to} 120 _-4 .
Referring to n, an image (picture) of a sub-image signal to be encoded by the processors 130 _{-1 to} 130 _-4 in the divided image encoding devices 120 _{-1 to} 120 _-4 based on the temporal change of the characteristic parameter Xi, n. Predict the feature parameter that indicates the complexity of
The divided image encoding device 12 with reference to the characteristic parameters
0 _-1 120 processor 130 in the _{-4 -1} to 130 _-4 performs the allocation of the bit rate of the sub-image signal when coding, dividing the image coding apparatus that the bit rate 120 _-1
And it sends to the processor 130 _-1 130 _-4 in 120 _-4.

FIG. 4 shows the allocation bit rate calculation unit 17
From the encoding result up to the n-th picture (frame) in No. 1, the processor 13 in the divided image encoding devices 120 _{-1 to} 120 _-4 determines the temporal change of the characteristic parameter Xi, n.
The characteristic parameter Xi, n of the (n + 2) th frame regarding the complexity in the picture to be coded from 0 _{-1 to} 130 _-4
Is a graph showing the state of prediction of the feature parameter Xi, n regarding the complexity of the (n + 2) th picture. Assuming that the prediction feature parameter is X'i, n + 2, the prediction feature parameter X'i, n + 2 can be obtained by the following formula 4.

[0048]

[Equation 22]     X'i, n + 2 = (Xi, n) +2 * [(Xi, n-Xi, n-1)]               = 3 · (Xi, n-2) -2 · (Xi, n-1)                 However, Xi, n is the sub-image signal i of the nth frame                                 Is a characteristic parameter,                         Xi, n-1 is the sub-image signal i of the (n-1) th frame                                 Is a characteristic parameter of.                                                         ... (4)

FIG. 5 shows how the characteristic parameters concerning the complexity of the (n + 1) th picture are predicted from the coding results up to the nth picture when the image coding apparatus has sufficient transfer capacity and processing capacity. It is a graph.
Also in the case illustrated in FIG. 5, assuming that the prediction value for the (n + 1) th picture is (X′i, n + 1), the prediction value (X′i, n + 1) can be calculated by the following equation 5. .

[0050]

[Equation 23]     X'i, n + 1 = Xi, n + (Xi, n-Xi, n-1)               = 2 * (Xi, n)-(Xi, n-1)                                                         ... (5)

Equation 5 shows a linear relationship, which is the feature parameter (Xi, n-1) of the pictures of the past and present frames.
), Xi, n, the next frame (n +
It is shown that the characteristic parameter of the picture of 1) can be predicted as (X'i, n + 1).

Consider the meanings of equations (4) and (5).
In an image obtained by photographing with a camera or the like, it can be considered that the feature amount related to the complexity of the image rarely changes suddenly and extremely. Therefore, in addition to the feature parameter Xi, n regarding the image complexity of the n-th picture, the feature parameter (Xi, n-1) regarding the image complexity of the previous (n-1) -th picture is referred to. Observing the amount of change in the characteristic parameters, the characteristic parameters (X'i, n + 2) and (X'i, n + 1) related to the complexity of the images of the (n + 2) th picture and the (n + 1) th picture are observed.
) Can be predicted. In this way, by predicting the characteristic parameter concerning the complexity of the image in the picture to be encoded next, obtaining the bit rate by referring to the predicted characteristic parameter, and encoding at that bit rate,
It is possible to allocate a bit rate to a moving image signal without a time delay. By using the bit rate of such allocation, the divided image encoding device 120
By encoding the sub-picture signal of the next frame in the processors 130 _{-1 to} 130 _-4 in the _{-1 to} 120 _-4 , high quality encoding can be performed without time delay.

The method of predicting the feature parameter in the allocation bit rate operation unit 171 is not limited to the above-mentioned example, and the feature parameter relating to the image complexity in the picture to be coded is determined from the temporal change of the feature parameter relating to the image complexity. Anything that predicts can be used.

The calculation for obtaining the allocation of the bit rate R from the characteristic parameter relating to the complexity of the predicted image in the allocation bit rate calculating section 171 can be performed, for example, according to the following equation 6. In the formula below, the symbols X'1, n + 1, X'2, n + 1, X'3, n + 1, X '
4, n + 1 is the predicted value of the characteristic parameters relating to the complexity of the image of the sub-image signals to be encoded in the processor 130 _-1 130 _-4 divided image encoding device 120 _-1 120 within _-4.

[0055]

[Equation 24]     Ri, n + 1 = Rx (X'i, n + 1)               / (X'1, n + 1 + X'2, n + 1 + X'3, n + 1 + X'4, n + 1)                                                         ... (6)

The allocation bit rate calculation unit 171 uses the equation 6
The allocation bit rate calculated based on the processor 130 _-1 in the divided image coding devices 120 _{-1 to} 120 _-4 .
Given to 130 _-4, processor 130 _-1 130 _-4 encode the next sub-image signals in a given bit rate.

Regarding the overall operation of the image encoding device 101
I will explain. For example, the luminance signal is 1920 pixels wide x 1 vertical
080 lines, color difference signal is 960 horizontal pixels x 1080 vertical pixels
In HDTV video signal is input to the video encoding device 101.
Then, in the image dividing device 110, the luminance signal is
1440 pixels in the horizontal direction and 720 pixels in the horizontal direction for the color difference signal.
Convert to plain. In the vertical direction, the luminance signal and color difference signal
Dummy data of 8 lines is attached under each image.
To be 1088 lines. And like this
The converted image signal has four areas A, as shown in FIG.
B. It is divided into C and D, and the first to fourth divided image codes are respectively
Encryption device 120 _-1~ 120_-FourIs output to.

The first to fourth divided image coding devices 120 _-1
120 _-4 Each processor 130 _-i of (i = 1 to 4) is
A sub-image signal of horizontal 720 × vertical 544 lines divided by MPEG2-MP @ ML is encoded based on a bit rate R obtained from a characteristic parameter relating to image complexity predicted by the allocation bit rate calculation unit 171. To do. Each processor 130 _-i (i = 1 to 4) uses the generated bit amount Si, n and the average quantization scale Qi, n obtained by the encoding for each picture, which is obtained by the encoding, as described above. The characteristic parameter Xi is obtained by the above method and is output to the allocation bit rate calculation unit 171 of the control device 170.

The video stream coded by each processor 130 _-i (i = 1 to 4) is temporarily stored in the buffer 150 _-i in each divided image coding device 120 _-i .

The allocation bit rate calculation unit 171 predicts the characteristic parameter by the above-described method, performs the allocation calculation of the bit rate R using the predicted characteristic parameter, and calculates the first to fourth divided image encoding devices 120. _{-1 to} 120
_-4 to each processor 130 _-i (i = 1 to 4).

FIG. 6 shows the first to fourth divided image coding devices 1.
20 _{-1 to} 120 _-4 of each processor 130 _-i (i = 1 to 1
4) The feature parameter Xi output from 4) and the feature parameter Xi in the allocation bit rate calculation unit 171 are input to predict the feature parameter, and the bit rate R is calculated using the predicted feature parameter. 1st-4th division | segmentation image coding apparatus 120 < _-1 > -120
₄ is a graph showing a timing of output to each processor 130 _-i (i = 1 to 4) of _-4 .

When the encoding of the n-th picture (sub-image signal) is completed in each processor 130 _-i (i = 1 to 4) of the divided image encoding devices 120 _{-1 to} 120 _-4 , the resulting image is obtained. Of the characteristic parameters related to the complexity of the bit rate calculation unit 171 in the external control device 170.
Transfer to. The allocation bit rate calculation unit 171 includes the processors 1 of the divided image coding devices 120 _{-1 to} 120 _-4.
In 30 _-i (i = 1 to 4), the bit rate to be assigned to each divided image encoding device is calculated during the period of encoding the (n + 1) th picture. The allocated bit rates are divided image coding devices 120 _{−1 to} 120
_-4 in each processor 130 _-i (i = 1 to 4) (n
Before the encoding of the +2) th picture is started, the calculated allocation bit rate is transferred. Each of the processors 130 _-i (i of the divided image coding devices 120 _{-1 to} 120 _-4
= 1 to 4), the (n + 2) th picture is encoded according to the transferred bit rate.

At such operation timing, (n +
The bit rate assigned when the 2) -th picture is encoded depends on the characteristic parameter relating to the image complexity of the n-th picture, and there is a delay of 2 pictures. Provisionally, the divided image coding devices 120 _{-1 to} 120 _-4
Even if the computing capacity of the external control device is sufficient, it is the (n + 1) th picture that reflects the characteristic parameter relating to the image complexity of the nth picture.
One picture delay occurs. Therefore, even if the characteristic parameter relating to the image complexity changes during this delay, it is not correctly reflected in the bit rate. However, in an image obtained by photographing with a camera or the like, it is considered that the characteristic parameter relating to the complexity of the image rarely changes suddenly extremely. Therefore, as described with reference to FIG. In addition to the feature parameter related to complexity, the amount of change of the feature parameter is observed from the feature parameter related to the complexity of the image of the previous picture, and the feature parameter related to the image complexity in the (n + 2) th and (n + 1) th pictures is determined. Predict and allocate a better bit rate. As a result, high quality encoding can be performed.

The first to fourth divided image coding devices 120 _-1
~ 120 _-4 , the allocation bit rate operation unit 1
Since the encoding is performed at the variable bit rate designated by 71, the buffer 150 _{-1 of} each of the divided image encoding devices A to D
The data occupancy of 150 _-4 changes as shown in FIG.

The image coding apparatus 101 time-divisionally multiplexes four video streams and transmits them as one video stream, and does not simply transmit the four video streams in parallel. Therefore, the change in the buffer occupancy as the sum is as illustrated in FIG.

The first to fourth divided image coding devices 120 _-1
~ 120 _-4 processors 130 _{-1 to} 130 _-4 are VB
Encoding processing is performed based on the occupied amount of the V buffer.

The first to fourth divided image coding devices 120 _-1
The video streams stored in the respective buffers 150 _-i in 120 to 120 ₋₄ are output to the video stream integration device 160 at the bit rate allocated by the allocation bit rate operation unit 171, and the video stream integration device 160 is output.
In, the video stream is combined into one MPEG2-MP @ HL video stream, that is, decomposed in slice units. As a result, as shown in FIG. 3, A1, B1, A2,
B2, ..., A34, B34, C1, D1, C2, D2
, C34, D34 are rearranged in this order and output from the video stream integration device 160.

The MPEG2-M generated in this way
The P @ HL video stream is multiplexed with an audio signal, if necessary, and then transmitted through a transmission path or recorded on a recording medium such as the video tape 300 shown in FIG. FIG. 9 shows a situation in which the video stream thus integrated is transmitted in the band of the transmission rate R.
As shown in FIG. 3, since the image signals are integrated in the line direction, they are transmitted in the order shown in the circle in FIG.

As described above, according to the present embodiment, image processing is performed in each processor 130 _-i (i = 1 to 4) of each of the first to fourth divided image coding devices 120 _{-1 to} 120 _-4 . A feature parameter relating to complexity is calculated, and the allocation bit rate operation unit 171 predicts a feature parameter relating to image complexity of a picture to be encoded next from the temporal change of the feature parameter relating to image complexity, and is suitable for that. Bit rate allocation is performed, and in each processor 130 _-i (i = 1 to 4) of the first to fourth divided image coding devices 120 _{-1 to} 120 _-4 , the next sub-image signal (picture ) Is encoded, so
High-quality encoding processing can be performed without time delay.

Note that a television image signal is usually accompanied by an audio signal. Audio signal encoding is performed by a device different from the above-described image signal encoding process,
It is encoded and recorded together with the image signal on, for example, a video tape.

The image decoding apparatus of the first embodiment will be described. FIG. 10 is a block diagram showing the configuration of the image decoding apparatus 201 according to the first embodiment. The image decoding device 201 includes a video stream separation device 210, first to fourth divided image decoding devices 220 _{-1 to} 220 _-4 , an image synthesizing device 260, and a control device 270. Since the image decoding device 201 performs a decoding process that is the reverse of the encoding process performed by the image encoding device 101 illustrated in FIG. 1, the configuration of the image decoding device 201 is basically the same as the image encoding device 101.
The configuration is reversed and the reverse processing is performed.

The image decoding apparatus 201 according to the preferred embodiment is the same as the image coding apparatus 101 described above.
A preferred embodiment is shown in the case of encoding a large amount of image signals such as an HDTV signal. In such a case, when one frame of the HDTV signal is encoded by one encoding means, the encoding is performed. Since it takes a very long time to process and it becomes difficult to realize the encoding process in real time, the case of encoding the sub-image signal obtained by dividing the image signal into N pieces has been exemplified. The case of decoding the converted image signal will be described. However, in principle, the present invention can be decoded without dividing the encoded image signal. Hereinafter, a case where the coded image signal is divided into N pieces and the decoding processing is performed will be described. In the present embodiment, the number N indicating a plurality of divisions is 4.

The video stream separation device 210 reads out the video stream coded by the image coding device 101 illustrated in FIG. 1 and recorded on the video tape 300, for example. If necessary, an audio signal may also be encoded by a device (not shown) and recorded on the video tape 300 together with the video stream. However, since the subject of the present invention is the encoding and decoding of image signals, audio signals are not mentioned. Video stream separation device 2
10 is one M read from the video tape 300.
4 PEG2-MP @ HL video streams
It is separated into PEG2-MP @ ML video streams and output to the first to fourth divided image decoding apparatuses 220 _{-1 to} 220 _-4 . The video stream separation device 210 basically decomposes one MPEG2-MP @ HL video stream in slice units, distributes them into four, and reconstructs them to thereby reconstruct the four MPEG2-MP @ ML. Get the video stream. Since the sequence higher than the slice unit, GOP, picture level header data, extension data, and user data are common, the same video stream is used for four.

More specifically, when the sequence start code appears in the input video stream, the video stream separation device 210 performs the sequence
The input video stream from the start code until the first slice start code appears
The divided image decoding devices 220 _{-1 to} 220 _-4 are output to all four. At the same time, an instruction from the control device 270 is issued regarding parameters that need to be rewritten to form four MPEG2-MP @ ML video streams, such as horizontal image size, vertical image size, aspect ratio information, VBV buffer size, and VBV delay. Rewrite based on.

A value at the time of encoding cannot be obtained for the VBV delay, but this value is "0x" in a general decoder.
Decoding is possible by setting FFFF ". Regarding the horizontal image size and vertical image size, four sets may be obtained from the input video stream, but since complicated calculations are required, it is fixed for the entire system. It is desirable to use a value or to notify it from the control device 270 in some way.The bit rate cannot be obtained at the time of encoding, but if it is set to a value larger than the original bit rate, a general decoder can use it. Since it can be decoded, it is output as it is without being rewritten.The bit rate value written in the video stream may be larger than the maximum bit rate of MP @ ML, but the actual bit rate value is , It is less than the maximum bit rate (because it is encoded that way when encoding). If the value of the bit rate MP @
It may be rewritten to the maximum bit rate of ML.

The video stream after the slice start code is decomposed in slice units, and the decomposed video stream in slice units is distributed to the first to fourth divided image decoding devices 220 _{-1 to} 220 _-4. Output. As shown in FIG. 3, the decomposed video streams of slice units are A1, B1, A2, B2 ,. A34, B3
4, C1, D1, C2, D2, ..., C34, D34 are arranged in this order. Therefore, for the first divided image decoding device 220 _-1 (divided image decoding device A), A1, A
2, ..., A34, second divided image decoding device 220
_-2 (divided image decoding device B), B1, B2
, B34, C1, C2, ..., C3 for the third divided image decoding device 220 _-3 (divided image decoding device C)
4, D4 is output to the fourth divided image decoding device 220 _-4 (divided image decoding device D) in the order of D1, D2, ..., D34.

At this time, the value of the slice header indicating the vertical position information of the slice is rewritten to a correct value as necessary. Also, the macroblock address, which represents the horizontal position information in the first macroblock of the slice,
The increment is also rewritten to the correct value if necessary.
If the sequence end code is input after the video stream of the slice D34 is output, the video stream separation device 210 outputs the first to fourth divided image decoding devices 2
The sequence end code is output to all 20 _{-1 to} 220 _-4 and the separation is completed or the next sequence start code is waited for. If the sequence end code is not output, the next picture start code or group start code (GOP start code)
Is output, the video stream separation device 210 repeats the same processing as above.

The first to fourth divided image decoding devices 220 _-1
220 _-4 buffer the input bitstreams, sequentially decode them, and output them to the screen synthesizing device 260. Each divided image decoding device 220 _-i (i = 1 to 4)
_Has a buffer 230 _-i and a processor 240 _-i . The buffer 230 _-i is a buffer for temporarily storing the input coded video stream, and is a timing at which the amount of data required for decoding is decoded from the buffer 230 _-i to the processor 240 _-i. Substantially instantaneous reading is performed in picture units.

The processor 240 _-i is an MPEG2-MP
The @ML decoder decodes the input coded video stream by MPEG2. For example, in the case of a video stream encoded by the image encoding device 101 according to the above-described embodiment, this decoding causes the luminance signal to be 720 horizontal pixels × 544 vertical lines in a 4: 2: 0 format with a frame rate of 30 frames. / Sec interlace signal is restored and output.

The image synthesizing device 260 synthesizes the four moving image signals output from the first to fourth divided image decoding devices 220 _{-1 to} 220 _-4 into one moving image signal to generate the original moving image. The signal is restored and output. Specifically, the image synthesizing device 260 includes the first to fourth divided image decoding devices 220.
_The four moving image signals output from _{-1 to} 220 _-4 are combined. As a result, the size of the luminance signal is 1440 pixels horizontally × 1088 lines vertically, and the size of the color difference signal is 720 pixels horizontally ×
With 544 lines in the vertical direction, a moving image signal of 30 frames / second can be obtained in the 4: 2: 0 format.

The image synthesizing device 260 removes the dummy data at the bottom of the screen attached in the image dividing device 110 of the image encoding device 101 as shown in FIG. 2, and the luminance signal has a size of 1440 horizontal pixels × 1080 vertical lines. The size of the color difference signal is 720 pixels horizontally × 540 lines vertically.
Next, the image synthesizing device 260 performs the reverse process of the filtering process performed by the image dividing device 110 of the image encoding device 101, and the size in the horizontal direction increases. The size of the luminance signal is 1920 pixels horizontally × 1080 lines vertically, and the size of the color difference signal is 960 pixels horizontally × 540 lines vertically. The moving image signal thus generated is output from the image decoding device 201 and displayed on, for example, a video monitor.

The overall operation of the image decoding apparatus 201 will be described. The video decoding device 201 receives the video tape 30 from the transmission line or as shown in FIG.
MPEG2 reproduced and input from a recording medium such as 0
In the MP @ HL video stream, after the audio signal is separated as necessary, the video stream separation device 210
Entered in. One video stream input to the image decoding device 201 is four MPEG2-MP @ ML.
Of the video stream and are input to the first to fourth divided image decoding devices 220 _{-1 to} 220 _-4 . That is, one MPEG2-MP @ HL video stream is decomposed in slice units, and these are divided into four to be reconstructed to obtain four MPEG2-MP @ ML video streams.

The video stream separation processing will be described in more detail.
In short, the video stream separation device 210 is
Sequence start code in the video stream
When it appears, the first start
Input video until the rice start code appears
The stream is divided into first to fourth divided image decoding devices 220. _-1
~ 220_-FourOutput to all of. At the same time, 4 MP
It is written to make it a video stream of EG2-MP @ ML.
Parameters that need to be changed, such as the horizontal image size
, Vertical image size, aspect ratio information, VBV buffer
From the controller 270.
Rewrite according to the instructions.

When the slice start code appears,
Subsequent video streams are decomposed in slice units, and A1 and A are given to the first divided image decoding device 220 _-1 .
2, ..., A34, B1, B2, ..., B34 for the second divided image decoding device 220 _-2 , and C1, C2, ..., for the third divided image decoding device 220 _-3 . C34, and D for the fourth split image decoding device 220 _-4
, D2, ..., D34 are output in this order.

The first to fourth divided image decoding devices 220 _-1
To 220 _-4 , the MPEG2-MP @ ML video streams are once input to the buffers 230 _-i (i = 1 to 1).
4) and then read by the processor 240 _-i and decoded. As a result, the luminance signal is 720 horizontal pixels ×
Vertical 544 lines, 4: 2: 0 format and frame rate 3
An interlace signal of 0 frame / second is output.

The first to fourth divided image decoding devices 220 _-1
All of the four moving image signals output from -220 _-4 are input to the image synthesizing device 260 and are first synthesized into one moving image signal. That is, the luminance signal is 1440 horizontal pixels × 1088 vertical lines, and the color difference signal is 720 horizontal pixels × 54 vertical pixels.
With 4 lines, a composite image of 4: 2: 0 format 30 frames / sec is generated. Next, the dummy data at the bottom of the screen is removed, the luminance signal is set to 1440 horizontal pixels × 1080 vertical lines, and the color difference signal is set to 720 horizontal pixels × 540 vertical lines, and then the image division device 110 of the image encoding device 101 performs the processing. The inverse processing of the filtered processing performed is performed, and the luminance signal is set to 192
0 pixels × 1080 lines vertically, color difference signal is 960 pixels horizontally ×
Converted to vertical 540 lines. The moving image signal generated in this manner is output from the screen synthesizing device 260 and input to, for example, a video monitor to be displayed.

As described above, in the image coding apparatus 101 and the image decoding apparatus 201 according to the present embodiment, FIG.
As described above, the screen is divided into four parts, each of which is encoded by a separate encoding device and then decoded by the decoding device. At that time, even if the bit rate is constant as a whole, the encoding apparatus encodes at a variable bit rate according to the picture, the movement, or the complexity of the image. Therefore, a high-quality image can be obtained even when the design or movement greatly differs depending on the complexity of the image or the position of the screen.

When carrying out the image coding apparatus, the image decoding apparatus, and these methods of the present invention, various modifications can be made without being limited to the above-mentioned embodiments.

For example, MPEG2 exemplified as the above-mentioned encoding and decoding system is merely an example. Depending on the type of image signal to be encoded and decoded, the degree of encoding, etc., an appropriate encoding method and decoding method existing or devised in the future can be adopted.

In the above-described embodiment, the image signal to be encoded is a television image signal, particularly,
Although an HDTV signal has been illustrated, in implementing the present invention,
The present invention is not limited to the HDTV signal and the television signal, but can be applied to various image signals.

As described above, it is necessary to divide and integrate the image signals in the image coding apparatus, but the division and integration of the coded image signals in the image decoding apparatus are not essential for implementing the present invention. The number of divisions is appropriately determined according to the size of the image signal to be encoded, the device scale of the image encoding device and the image decoding device, and the processing speed.

[0092]

According to the image coding apparatus and the image coding method of the present invention, the characteristic parameter relating to the image complexity is calculated, and the coding is performed next from the temporal change of the characteristic parameter relating to the image complexity. It predicts the characteristic parameter related to the complexity of the picture image, allocates the bit rate suitable for it, and encodes the next image signal (picture) at that bit rate. It can be performed.

According to the image decoding apparatus and the image decoding method of the present invention, since the image signal coded by the above method is decoded, it is possible to perform a high quality decoding process without a time delay.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an image coding apparatus according to a first embodiment of the present invention.

FIG. 2 is a graph illustrating a configuration of an image signal and an HDTV video signal encoded by the image encoding device shown in FIG. 1 and a division / division method thereof.

FIG. 3 is a diagram illustrating the arrangement of video streams decomposed into slices in the image encoding device in FIG. 1.

FIG. 4 is a diagram showing the results of encoding up to the n-th picture in the allocation bit rate operation unit illustrated in FIG. 1, based on the temporal change of the characteristic parameter Xi, n, which is encoded by the processor in the divided image encoding device. 7 is a graph showing a state of predicting a feature parameter Xi, n related to complexity in a picture to be converted and predicting a feature parameter Xi, n related to complexity of an (n + 2) th picture.

FIG. 5 is a characteristic parameter regarding the complexity of the (n + 1) th picture from the coding result up to the nth picture when the image coding apparatus illustrated in FIG. 1 has sufficient transfer capacity and processing capacity. It is a graph showing how to predict.

FIG. 6 is a characteristic parameter Xi output from each processor in the divided image encoding device illustrated in FIG. 1 and a characteristic parameter Xi in an allocation bit rate calculation unit.
To predict the characteristic parameter, and using the predicted characteristic parameter, perform bit rate allocation calculation,
It is a graph which shows the timing output to each processor.

7 is a diagram showing an example of the data occupancy of each buffer of the four divided image encoding devices of the image encoding device shown in FIG. 1.

8 is a diagram showing a data occupancy amount as a total sum of the buffers shown in FIG. 7;

9 is a diagram showing a situation in which a video stream integrated by the video stream integration device of the image encoding device shown in FIG. 1 is transmitted in a band of a predetermined transmission rate.

FIG. 10 is a block diagram showing a configuration of an image decoding apparatus according to the first embodiment of the present invention.

[Explanation of symbols]

101 ... Image encoding device, 110 ... Image dividing device, 12
0 ... Division image coding device, 130 ... Processor, 150
... buffer, 160 ... video stream integration device, 17
0 ... Control device, 171 ... Allocation bit rate calculation unit,
172 ... Control unit, 201 ... Image decoding device, 210 ... Video stream separation device, 220 ... Divided image decoding device, 230 ... Buffer, 240 ... Processor, 260 ...
Image synthesizer 270 ... Control device

Claims (22)

[Claims] 1. An image signal dividing means for dividing a two-dimensional image signal into N two-dimensional sub-image signals, and a plurality of the two-dimensional N-divided sub-image signals are designated respectively. N encoding means for encoding at a bit rate, characteristic parameter calculating means for calculating a characteristic parameter indicating the complexity of the image of the sub-image signal from the encoded result, and the N calculated Allocating a bit rate to be used in each of the N encoding means from the characteristic parameter,
And a coded image signal unifying unit for unifying the N sub-image signals coded by the N coding units, wherein each of the N coding units includes: The next divided two-dimensional sub-image signal is encoded according to the bit rate calculated by the allocation bit rate calculation means, and the characteristic parameter calculation means calculates the complexity of each image of the two-dimensional sub-image signal from the encoded result. An image encoding apparatus that calculates a characteristic parameter indicating the degree. 2. The encoding means is a characteristic parameter Xi, n (i = 1, ..., N, which indicates the complexity of the image, and n
Is the n-th frame) is calculated based on the following equation. ## EQU1 ## Xi, n = Si, n × Qi, n where Si, n is the generated code bit amount of n frames of the sub-image signal i, and Qi, n is the average quantum of n frames of the sub-image signal i. It is a scale. 3. The encoding means is a characteristic parameter Xi, n (i = 1, ..., N, which indicates the complexity of the image, and n
Is the n-th frame) is calculated based on the following equation. ## EQU00002 ## Xi, n = .SIGMA. (Si, j.times.Qi, k) where Si, j is the generated code bit amount for each j frame of the sub-image signal i (j = nl, ..., n). -1, n) Qi, k is an average quantization scale for each k frame of the sub-image signal i (k = nm, ..., N-1, n). 4. The encoding means is a characteristic parameter Xi, n (i = 1, ..., N, which indicates the complexity of the image, and n
Is the n-th frame) is calculated based on the following equation. ## EQU3 ## Xi, n = .SIGMA. (Si, j * Qi, k) / (N-1) where Si, j is the generated code bit amount of the j-frame image of the sub-image signal i (j = nl, ..., n-1, n) Qi, k is an average quantization scale of k frames of the sub-image signal i (k = nm, ..., n-1, n). 5. The allocated bit rate calculation means calculates a predicted value of a characteristic parameter relating to the complexity of an image of a sub-image signal to be encoded next from the temporal change of the characteristic parameter, and the calculated predicted characteristic parameter The image coding apparatus according to claim 1, wherein the allocated bit rate is calculated from 6. The allocation bit rate calculation means,
The image coding apparatus according to claim 5, wherein the predicted value (X'i, n + 2) of the feature parameter of the (n + 2) th frame is obtained by the following formula. ## EQU00004 ## X'i, n + 2 = (Xi, n) +2 * [(Xi, n-Xi, n-1)] = 3 * (Xi, n-2) -2 * (Xi, n- 1) where Xi, n is a characteristic parameter of the sub-image signal i of the nth frame, and Xi, n-1 is a characteristic parameter of the sub-image signal i of the (n-1) th frame. 7. The allocation bit rate calculating means,
The image coding apparatus according to claim 5, wherein the predicted value (X'i, n + 1) of the characteristic parameter of the (n + 1) th frame is obtained by the following formula. X'i, n + 1 = Xi, n + (Xi, n-Xi, n-1) = 2 * (Xi, n)-(Xi, n-1) where Xi, n is n Xi, n-1 is a feature parameter of the sub-image signal i of the (n-1) th frame. 8. The allocated bit rate calculating means is a prediction feature parameter (X'i, n +) relating to the complexity of the image.
The image coding device according to claim 6 or 7, wherein the allocation of the bit rate R from 1) is calculated based on the following formula. ## EQU00006 ## Ri, n + 1 = R.times. (X'i, n + 1) /. SIGMA. (X'i, n + 1) 9. The image coding apparatus according to claim 1, wherein each of the N coding means has a buffer for storing a video stream generated by coding. 10. Each of the N encoders has a VBV
10. A video buffering verifier buffer is provided, and each of the N encoding devices performs the encoding based on an occupied amount of the VBV buffer provided therein.
The image encoding device according to 1. 11. An image signal dividing step of dividing a two-dimensional image signal into N two-dimensional sub-image signals, and a bit rate for specifying the N divided two-dimensional sub-image signals. And an encoding step of calculating the characteristic parameter indicating the complexity of the image of each of the two-dimensional sub-image signals from the encoding result, and a characteristic parameter calculating step of calculating the characteristic parameter from the calculated characteristic parameter. An allocation bit rate calculating step of calculating a bit rate in the encoding step and reflecting the calculated bit rate in the processing of the encoding step; and a two-dimensional sub-image signal encoded in the encoding step. A coded image signal integrating step of integrating, wherein in the encoding step, the next plurality of divided secondary signals according to the bit rate calculated in the allocation bit rate calculating step are performed. The sub-picture signal coded, calculates a characteristic parameter indicating the complexity of the feature parameters in the calculation step the result of the encoding of two-dimensional form sub-image signals image, the image coding method. 12. In the encoding step, a characteristic parameter Xi, n (i = 1, ..., N, where n represents an nth frame) indicating the complexity of the image is calculated based on the following equation. The image coding method according to claim 11, which is calculated. Xi, n = Si, n × Qi, n where Si, n is the generated code bit amount of n frames of the sub-image signal i, and Qi, n is the average quantum of n frames of the sub-image signal i. It is a scale. 13. In the encoding step, a characteristic parameter Xi, n (i = 1, ..., N, where n represents the nth frame) indicating the complexity of the image is calculated based on the following equation. The image coding method according to claim 11, which is calculated. Xi, n = Σ (Si, j × Qi, k) where Si, j is the generated code bit amount of j frameme of the sub-image signal i (j = nl, ..., n- 1, n) Qi, k is an average quantization scale of k frames of the sub-image signal i (k = nm, ..., N-1, n). 14. In the encoding step, a characteristic parameter Xi, n (where i = 1, ..., N, where n represents the nth frame) indicating the complexity of the image is calculated based on the following equation. The image coding method according to claim 11, which is calculated. Xi, n = Σ (Si, j × Qi, k) / (N-1) where Si, j is the generated code bit amount of the j frame of the sub-image signal i (j = nl ,. .., n-1, n) Qi, k is the average quantization scale of k frames of the sub-image signal i (k = nm, ..., N-1, n). 15. A predictive value of a feature parameter relating to image complexity of a sub-image signal to be coded next is calculated from the temporal change of the feature parameter in the allocation bit rate calculation step, and the calculated predictive feature parameter The image encoding method according to claim 11, wherein the allocated bit rate is calculated from 16. The predicted value (X ′ of the (n + 2) th feature parameter in the allocation bit rate calculation step.
The image coding method according to claim 15, wherein i, n + 2) is calculated by the following formula. X'i, n + 2 = (Xi, n) +2 * [(Xi, n-Xi, n-1)] = 3 * (Xi, n-2) -2 * (Xi, n- 1) where Xi, n is a characteristic parameter of the sub-image signal i of the nth frame, and Xi, n-1 is a characteristic parameter of the sub-image signal i of the (n-1) th frame. 17. The predicted value (X ′ of the (n + 1) th feature parameter in the allocation bit rate calculation step.
16. The image coding method according to claim 15, wherein i, n + 1) is calculated by the following formula. X'i, n + 1 = Xi, n + (Xi, n-Xi, n-1) = 2 * (Xi, n)-(Xi, n-1) where Xi, n is n Xi, n-1 is a feature parameter of the sub-image signal i of the (n-1) th frame. 18. The allocation of the bit rate R from the prediction feature parameter (X'i, n + 1) relating to the complexity of the image is calculated in the allocation bit rate calculation step based on the following equation. 16. The image coding method according to 16 or 17. ## EQU12 ## Ri, n + 1 = R × (X'i, n + 1) / Σ (X'i, n + 1) 19. The image coding method according to claim 11, wherein in the N coding steps, a video stream generated by coding is buffered. 20. The buffering is performed by VBV (Video B).
The image coding method according to claim 19, wherein the image coding method is uffering verifier) buffering. 21. A two-dimensional image signal is divided into N two-dimensional sub-image signals, the N divided two-dimensional sub-image signals are encoded at a designated bit rate, and the code is obtained. A characteristic parameter indicating the complexity of the image of each two-dimensional sub-image signal is calculated from the converted result, a bit rate used in the encoding process is calculated from the calculated characteristic parameter, and the encoding process is performed in the encoding process. An image decoding apparatus comprising: a decoding unit that receives an image signal obtained by integrating encoded two-dimensional sub-image signals and decodes the received image signal according to the calculated bit rate. 22. A two-dimensional image signal is divided into N two-dimensional sub-image signals, the N divided two-dimensional sub-image signals are encoded at a designated bit rate, and the code A characteristic parameter indicating the complexity of the image of each two-dimensional sub-image signal is calculated from the converted result, a bit rate used in the encoding process is calculated from the calculated characteristic parameter, and the encoding process is performed in the encoding process. An image decoding method comprising a decoding step of receiving an image signal obtained by integrating encoded two-dimensional sub-image signals and decoding the received image signal according to the calculated bit rate.