JPH0918879A - Moving image encoding method and moving image communication device - Google Patents

Abstract

PURPOSE: To suppress the unnecessary codes which are produced at the static part of an image and to improve the compression efficiency in a moving image encoding mode by changing the noise component generated at the static part of the image into zero based on the set threshold value in an INTER mode. CONSTITUTION: A comparator 13 compares the image data on the n-th frame stored in a 1st frame memory 12 with the input image data on the (n+1)-th frame and takes out the difference between both data. A deciding device 14 compares the difference with the threshold value set at a CPU to output the difference after changing into zero when the difference is smaller than the threshold value. Then the device 14 outputs the difference as it is when the difference is larger than the threshold value. The difference is added to the image data on the n-th frame by an adder 15, written in a 2nd frame memory 16, and compressed and encoded by a moving image encoding circuit 17. In such a constitution, the noises are erased at the static part of an image and the compression efficiency is improved in a moving image encoding mode.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a moving picture coding method for coding moving pictures in image communication and the like, and a moving picture communication apparatus for communicating moving pictures.

[0002]

2. Description of the Related Art In recent years, digital communication networks have become widespread, and
Along with this, video conference systems have become widespread.
It is considered that this tendency will continue to expand, and moving image coding technology is essential for moving image communication, which is the center of this trend.

A conventional moving picture coding method and a moving picture communication apparatus to which the method is applied will be described below with reference to the drawings.

FIG. 3 is a block diagram showing a moving picture communication apparatus to which a conventional moving picture coding method is applied. The device of FIG.
The H.264 standard, which is a standard method for moving picture coding of ITU-T. The video communication device to which the one cited from the H.261 recommendation is applied.
In FIG. 3, 1 is a subtracter for subtracting the image data input from the input terminal T1, 2 is a first switch for switching the data path between intraframe coding and interframe coding, and 2a is a common switch 2. Terminals 2b and 2c are switching terminals of the switch 2 and 3 is a discrete cosine transform (Dis-crete Co
DCT that performs sine transform (hereinafter referred to as "DCT")
Transformer 4, quantizer for quantizing, 5 inverse quantizer for inverse quantizing quantized data, 6 inverse discrete cosine transform (hereinafter referred to as "inverse DCT") ) Inverse DCT converter, 7 is 2
An adder for adding input data, 8 a motion compensation image memory having a variable delay function for motion compensation, 9 an in-loop filter as a low-pass filter for reducing the accumulation of errors due to inter-frame prediction, and 10 an adder 7, a loop composed of the image memory 8 for motion compensation and the in-loop filter 9 is turned on,
Second switch for switching off, 10a is switch 1
0 is a common terminal, 10b and 10c are switching terminals of the switch 10, and 11 is an encoding control unit for controlling the whole.

The operation of the moving picture communication device configured as described above will be described. There are two types of moving picture coding, intraframe coding and interframe coding. The coding mode for intraframe coding is called INTRA mode, and the coding mode for interframe coding is called INTER mode. First, the INTRA mode will be described. In this mode, the common terminal 2 of the first switch 2 and the second switch 10 is
a and 10a are connected to the upper terminals 2b and 10b. The image data input from the input terminal T1 is DCT-converted by the DCT converter 3 through the first switch 2 into macroblock units of 16 × 16 pixels. The DCT transform is a kind of frequency transform. The DCT-transformed image data is quantized by the quantizer 4, and the quantized output index q of the transform coefficient in the DCT transform is output. At the same time, the encoding control unit 11 sends INTER mode and INTRA.
The INTER / INTRA identification flag p for identifying the mode, the transmission / non-transmission identifying flag t for identifying the transmission and non-transmission, and the designation data qz for designating the quantization characteristic are output. The flags p and t and the designated data qz are information necessary for decoding. The INTRA mode is such that encoding is performed within one frame.

Next, the INTER mode will be described. In this mode, the common terminals 2a and 10a of the first switch 2 and the second switch 10 are the lower terminals 2c,
10c (FIG. 3 shows a connection state in the INTER mode). In this mode, first the above INTR
After performing the same operation as in the A mode, the quantized output index q output in the INTRA mode is converted to the inverse quantizer 5,
Decoding is performed by the inverse DCT converter 6, and the decoded data is input to the motion compensation image memory 8 through the adder 7. The motion compensation image memory 8 compares the image data of the next frame input from the input terminal T1 with the decoded data input via the adder 7, and generates the motion vector v. The result of comparing the image data and the decoded data over the two frames is input to the subtracter 1 as a motion component of the image through the in-loop filter 9 and is again DCT.
Encoding is performed by the converter 3 and the quantizer 4, and the quantized output index q is output. At the same time, as in the INTRA mode, the INTER / INTRA identification flag p, the transmission / non-transmission identification flag t, and the designated data qz designating the quantization characteristic are output, and the motion vector v and the in-loop filter 9 are set for each macroblock. An on / off signal f, which is a determination signal as to whether or not it has acted, is also output.

The above is H.264, which is a standard method of moving picture coding. 2
61 is an outline of operation in the 61-recommended system.

[0008]

However, in the moving picture communication device to which the above-mentioned conventional moving picture coding system is applied,
In the case of INTRA mode in which encoding is performed in one frame, since the difference data is not processed, there is no influence of image movement or noise, but in the case of INTER mode in which encoding is performed between frames, Since the method of comparing the image data of the preceding and succeeding frames with the decoded data and encoding the difference data, there is a change in the image data due to noise etc. in the static part of the image where no code originally occurs. However, there is a problem that an extra code is generated based on the change.

The present invention solves the above-mentioned conventional problems. Even if noise occurs in a still portion of an image in the INTER mode, the generated noise can be eliminated to improve the compression efficiency. Another object of the present invention is to provide a moving image communication device capable of eliminating the generated noise and improving the compression efficiency even if noise occurs in the still portion of the image in the INTER mode.

[0010]

In order to achieve this object, a moving picture coding method according to claim 1 of the present invention is applied to moving picture data of an nth (arbitrary natural number) frame of consecutive frames. A change amount calculating step of calculating a moving image data change amount as a level change amount of the moving image data of the (n + 1) th frame, and a level when the calculated moving image data change amount is smaller than a preset threshold value. Change the moving image data as a change to zero,
When the calculated moving image data change amount is larger than a preset threshold value, a determination step of setting the moving image data change amount as the level change amount to the calculated moving image data change amount, and a moving image as the level change amount Data change and n
It has a configuration including an addition step of adding moving image data of the th frame and an encoding step of compressing and encoding the added moving image data to generate encoded data.

According to a second aspect of the present invention, there is provided a moving image communication apparatus, wherein a first frame memory stores input moving image data in units of one frame, and n stored in the first frame memory.
The moving image data of the (arbitrary natural number) th frame and (n
The moving image data of the +1) th frame is compared in a fixed block unit, and the moving image data change amount as the level change amount of the moving image data of the (n + 1) th frame is compared with the moving image data of the nth frame. When the calculated moving image data change amount is smaller than a preset threshold value, the moving image data change amount as the level change amount is set to zero, and the calculated moving image data change amount is set in advance. And a moving image data change amount as the level change amount and the moving image data of the n-th frame. An addition unit for adding the data and outputting the addition moving image data; and a second unit for storing one frame of the addition moving image data output from the addition unit. A frame memory, a moving picture coding circuit for compressing and coding the added moving image data stored in the second frame memory and outputting the coded data, and communication for transmitting the coded data to the other party via the communication network. Circuit, a moving picture decoding circuit for decoding the encoded data of the other side from the communication circuit into the original added moving picture data, and the added moving picture data from the moving picture decoding circuit are stored as moving picture data in units of one frame. And a decoding frame memory for

[0012]

With the above construction, in the INTER mode in which the image data of the preceding and succeeding frames are compared and the difference data of these data is encoded, the noise component generated in the still portion of the image is set to zero by the set threshold value. Therefore, it is possible to suppress the generation of extra codes in the still portion of the image, and it is possible to improve the compression efficiency at the time of moving image coding.

[0013]

【Example】

(Embodiment 1) An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a moving picture communication apparatus to which a moving picture coding method according to an embodiment of the present invention is applied. In FIG. 1, reference numeral 12 denotes a first frame memory that stores one frame of moving image data input from the input terminal T1, and 13 denotes an n (arbitrary natural number) th frame stored in the first frame memory 12. Of the moving image data of the (n + 1) th frame and the moving image data of the (n + 1) th frame are compared in a fixed block unit, and the level difference of the moving image data of the (n + 1) th frame is compared with the moving image data of the nth frame. A comparator (comparison means) for calculating the moving image data change amount, 14 sets the moving image data change amount to be output to zero when the moving image data change amount calculated by the comparator 13 is smaller than a preset threshold value. , Comparator 1
When the moving image data change amount calculated in 3 is larger than a preset threshold value, the moving image data change amount to be output is used as the moving image data change amount calculated by the comparator 13 (judging means), 15 Is an adder (adding means) that adds the moving image data change output from the determiner 14 and the moving image data of the nth frame to output the added moving image data, and 16 is output from the adder 15. A second frame memory for storing one frame of added moving image data, 1
Reference numeral 7 is a moving picture coding circuit which compresses and codes the added moving image data stored in the second frame memory 16 and outputs the coded data, and 18 is a CPU which sets the threshold value in the judging device 14 in advance. .

The operation of the moving picture communication apparatus configured as described above will be described below. First, one frame of moving image data input from the input terminal T1 is stored in the first frame memory 12. Since the moving image data is continuously input from the input terminal T1, the image data of the nth frame stored in the first frame memory 12 and
By inputting the image data of the next (n + 1) th frame to the comparator 13 and comparing the two, the difference between the frames (image data change amount), that is, the amount of movement data can be extracted (change). Minute calculation step). In this case, in the moving image encoding, the image data is encoded in a fixed block (hereinafter referred to as “macroblock”) unit such as 16 × 16 pixels, and therefore the comparator 13 also compares the image data in macroblock units.

The decision unit 14 converts the above-mentioned motion amount data output from the comparator 13 as follows. First, the threshold value set in advance by the CPU 18 is compared with the motion amount data output from the comparator 13. As a result, when the motion amount data (moving image data change amount) is smaller than the threshold value, the macro block is a still part of the image, and the motion amount data output from the comparator 13 changes due to noise. Then, the decision unit 14 sets the motion amount data to zero and outputs it (decision step). Similarly, when the amount of motion data is larger than the threshold value, the macroblock is a moving part of the image, and the amount of motion data output from the comparator 13 is regarded as the original amount of motion of the image. The device 14 outputs the movement amount data as it is without changing it (decision step). In this way, the judging device 14 judges a still block and a moving block,
motion amount data (moving image data change amount) between n frame moving image data and (n + 1) frame moving image data;
The adder 15 adds n frames of moving image data from the first frame memory 12 (adding step).
Output data from this adder 15 (added moving image data)
In the case of a macro block in the still part of the image, the moving image data is exactly the same as in the n frame, and in the case of a macro block in a moving part, it is the moving image data in which the original motion is reflected. The image data is written in the second frame memory 16. That is, in the case of a macro block of the still portion of the image, n frame image data and zero data are written in the second frame memory 16,
In the case of a macro block of a moving portion, n frame image data and difference data (moving image data change amount) are written. The added moving image data stored in the second frame memory 16 is compressed and encoded by the moving image encoding circuit 17 and output as encoded data (encoding step).

As described above, according to the present embodiment, a macroblock of an image still portion, which is originally still and should have a difference of zero even when the preceding and following frames are compared, is quantized during A / D conversion. Even if noise is generated due to noise or the like and a certain difference value is generated, the difference value can be set to zero by a preset threshold value. Therefore, even in the INTER mode, an extra code of the still portion of the image is generated. Can be suppressed, unnecessary compression of extra codes is unnecessary, and compression efficiency at the time of moving image coding can be improved.

(Second Embodiment) A second embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a block diagram showing a moving picture communication device according to the second embodiment of the present invention. In FIG. 2, 12 is the first frame memory, and 13
Is a comparator, 14 is a determiner, 15 is an adder, 16 is a second frame memory, 17 is a moving picture coding circuit, 18 is a CPU
Since these are the same as those in FIG. 1, the same reference numerals are given and description thereof is omitted. Reference numeral 19 is an A / D converter for A / D converting the analog moving image signal input from the input terminal T1 to convert it into digital moving image data, and 20 is connected to the communication network, and the output data of the moving image encoding circuit 17 (code Communication data) to a partner video communication device or to receive coded data from the partner video communication device and output it to a video decoding circuit 21, which will be described later. Reference numeral 21 denotes a communication network from the partner video communication device. A moving picture decoding circuit that decodes the coded data that is input via the original added moving picture data, and a decoding unit 22 that stores the added moving picture data from the moving picture decoding circuit 21 as moving picture data in units of one frame. A frame memory 23 is a D / A converter which converts the digital moving image data output from the decoding frame memory 22 into an analog moving image signal and outputs the analog moving image signal from the output terminal T2. That.

The operation of the moving picture communication device configured as described above will be described. An image input device such as a video camera is generally connected to the input terminal T1, and the input analog moving image signal is converted into digital moving image data by the A / D converter 19. Hereinafter, as described in the first embodiment, the output data (n-frame moving image data) of the first frame memory 12 and the A / D converter 19 will be described.
Output data ((n + 1) frame moving image data) is compared by the comparator 13 on a macroblock basis, and the motion amount data is compared with a threshold value preset from the CPU 18 by the determiner 14. According to the magnitude, it is determined whether the block is a motion block or a still block, the change amount of the still block is corrected to zero and output, and the change amount of the motion block is output to the adder 15 as it is, and the determination unit 14 outputs the difference value. Output data (difference data) and the first frame memory 12
And the moving image data obtained by adding the moving image data of No. 2 to the second frame memory 16 and the moving image encoding circuit 17 compresses, encodes and encodes the added moving image data. Get the data. By this operation, the change amount of the moving image data of the still block is set to zero, so that the code generation in the INTER mode due to noise or the like can be set to zero. The above operation is similar to that of the first embodiment.

Next, the encoded data output from the moving image encoding circuit 17 is input to the communication circuit 20. Communication circuit 2
Reference numeral 0 is a circuit for connecting to a general communication network, and has a different structure depending on the type of communication network to be connected. For example, if the communication network is an analog line, the communication circuit 20 will be composed of a modem, and if it is an ISDN line, S /
If it is a T interface circuit or a LAN, it will be composed of a LAN controller circuit. In any case, the communication circuit 20 is connected to the video communication device of the other party and transmits / receives data to / from the other device. The encoded data output from the moving image encoding circuit 17 is transmitted to the partner device by the communication circuit 20, and the encoded data received from the partner device is decoded by the moving image decoding circuit 21. Video decoding circuit 2
Reference numeral 1 decodes the encoded data received from the partner device and converts it into added moving image data. Similar to the encoding, the decoding of the moving image is executed in macro block units, and the result is written in the decoding frame memory 22 as digital moving image data. The digital moving image data written in the decoding frame memory 22 is converted into an analog moving image signal by the D / A converter 23 and output. D / A
The analog moving image signal output from the converter 23 can be displayed on a television monitor or the like. In order to control the above operation, the CPU 18 sets the threshold value of the judging device 14, the operation control of the moving picture coding circuit 17, the connection / disconnection control of the communication circuit 20 with the partner device, the moving picture decoding circuit 2
1 operation control is performed.

A voice input device such as a microphone, a voice output device such as a speaker, a voice encoding / decoding circuit to which the respective voice input / output devices are connected to the moving image communication device, image encoded data and voice. It goes without saying that a videophone terminal device using images and voices can be configured by adding a demultiplexing circuit for demultiplexing encoded data and connecting to a communication circuit.

As described above, according to this embodiment, even if a certain difference value is generated due to noise generated in the macroblock of the still image portion due to quantization noise during A / D conversion, the CPU
Since the difference value can be set to zero by the threshold value preset in the judging device 14 by 18,
In the ER mode as well, it is possible to suppress the generation of extra codes in the still portion of the image, and it is possible to improve the compression efficiency during video encoding.

[0023]

As described above, according to the present invention, the moving image data of the (n + 1) th frame of the moving image data for the nth (arbitrary natural number) frame of the continuous frames is changed. A change amount calculation step of calculating the image data change amount, and when the calculated moving image data change amount is smaller than a preset threshold value, the moving image data change amount as the level change amount is set to zero, and the calculated moving image When the data change amount is larger than a preset threshold value, the moving image data change amount as the level change amount is determined as the calculated moving image data change amount, and the moving image data change amount as the level change amount is determined. An addition step of adding the moving image data of the n-th frame, and an encoding of generating the encoded data by compressing and encoding the added moving image data. By having a step,
In the INTER mode in which the image data of the preceding and succeeding frames are compared and the difference data of those data is encoded, the noise component generated in the still portion of the image can be made zero by the preset threshold value. It is possible to realize a moving image coding method that can suppress the generation of extra codes in the still portion of an image and can improve the compression efficiency during moving image coding.

Also, the first frame memory for storing the input moving image data in units of one frame, the moving image data of the n (arbitrary natural number) -th frame stored in the first frame memory, and (n + 1) ) The moving image data of the (n + 1) th frame is compared with the moving image data of the (n + 1) th frame to calculate the moving image data change amount as the level change amount of the (n + 1) th frame moving image data. When the calculated moving image data change amount is smaller than a preset threshold value, the moving image data change amount as the level change amount is set to zero, and the calculated moving image data change amount is set in advance. When it is larger than the threshold value, the moving image data change amount as the level change amount is determined as the calculated moving image data change amount, and the level change amount Adding means for adding the moving image data change amount and the moving image data of the n-th frame to output the added moving image data; and storing one frame of the added moving image data output from the adding means. No. 2 frame memory, a moving picture coding circuit for compressing and coding the added moving image data stored in the second frame memory and outputting the coded data, and sending the coded data to the other party via the communication network. Communication circuit, a moving picture decoding circuit that decodes the encoded data of the other side from the communication circuit into the original added moving picture data, and the added moving picture data from the moving picture decoding circuit in units of moving picture data By providing a decoding frame memory for storing as, it is possible to judge in the INTER mode in which the image data of the preceding and following frames are compared and the difference data of those data is encoded. The stage can set the noise component generated in the still portion of the image to zero by a preset threshold value, so that it is possible to suppress the generation of extra codes in the still portion of the image. It is possible to realize a moving image communication device that improves the compression efficiency in the above and can perform highly reliable communication with high communication efficiency.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a moving picture communication device to which a moving picture coding method according to an embodiment of the present invention is applied.

FIG. 2 is a block diagram showing a moving picture communication device according to a second embodiment of the present invention.

FIG. 3 is a block diagram showing a moving picture communication device to which a conventional moving picture coding method is applied.

[Explanation of symbols]

 12 First Frame Memory 13 Comparator (Comparison Means) 14 Judgment Machine (Judgment Means) 15 Adder (Addition Means) 16 Second Frame Memory 17 Video Coding Circuit 18 CPU 19 A / D Converter 20 Communication Circuit 21 Video Decoding circuit 22 Decoding frame memory 23 D / A converter T1 input terminal T2 output terminal

Claims (2)

[Claims] 1. (n + 1) for moving image data of an n-th (arbitrary natural number) frame of consecutive frames
A change amount calculating step of calculating a change amount of moving image data as a level change amount of moving image data of the th frame,
When the calculated moving image data change amount is smaller than a preset threshold value, the moving image data change amount as the level change amount is set to zero, and the calculated moving image data change amount is set in advance. When it is larger than the value, a step of determining that the moving image data change amount as the level change amount is the calculated moving image data change amount, the moving image data change amount as the level change amount and the moving image of the n-th frame A moving image coding method comprising: an addition step of adding image data, and an encoding step of compressing and encoding the added moving image data to generate encoded data. 2. A first frame memory for storing input moving image data in units of one frame, and moving image data of an n (arbitrary natural number) th frame stored in the first frame memory ( The moving image data of the (n + 1) th frame is compared in a fixed block unit, and the moving image data change amount as the level change amount of the moving image data of the (n + 1) th frame is compared with the moving image data of the nth frame. When the calculated moving image data change amount is smaller than a preset threshold value, the moving image data change amount as the level change amount is set to zero, and the calculated moving image data change amount is Judging means for setting the moving image data change amount as the level change amount to be the calculated moving image data change amount when it is larger than a preset threshold value An addition unit that adds the moving image data change amount as the level change amount and the moving image data of the n-th frame to output the added moving image data; and the addition moving image data output from the adding unit. A second frame memory for storing one frame, a moving picture coding circuit for compressing and coding the added moving image data stored in the second frame memory and outputting the coded data, and the coded data A communication circuit for transmitting to the other party via a communication network, a moving picture decoding circuit for decoding the other party's encoded data from the communication circuit into the original addition moving image data, and an addition from the moving picture decoding circuit A moving picture communication device, comprising: a decoding frame memory that stores moving image data as moving image data in units of one frame.