JPH07222173A - Picture processor - Google Patents

Abstract

PURPOSE:To reduce transmission data and to reproduce smooth pictures even in the communication network of a low speed rate at the time of conipressing and transmitting picture data. CONSTITUTION:The input part of the picture data is provided with a movement comparison part 1, the picture data of a previous frame and a present frame are compared and the difference data are computed. The output is supplied to a still judgement part 2 and the presence/absence of the change of the present frame against the previous frame are discriminated. When the change of pictures is present, the picture data are supplied to an intra-frame/inter-frame judgement part 4, a picture compression processing is performed and compressed data are transmitted through the communication network to a reception side. Also, when no change of the pictures is present, the identification signal is transmitted through a network interface processing part 10 to the reception side. Thus, the need of data transmission is eliminated during the time of a frame cycle without the change of the pictures and a transmission data amount is reduced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for compressing and transmitting image data.

[0002]

2. Description of the Related Art In recent years, CCITT (International Telegraph and Telephone Advisory Committee) has officially made a recommendation of video coding method, still image coding method, multiplexing method, communication procedure, etc. for audiovisual services, and accordingly, each company. To CCI
An image processing device such as a video conference system and a video phone that complies with the TT recommendation has been announced.

An example of a conventional image processing apparatus will be described with reference to FIG. FIG. 2 is a block diagram showing the overall configuration of a conventional image processing apparatus. In the figure, an image input (input image data) is input to the intra-frame / inter-frame determination unit 4 on a frame-by-frame basis. The intra-frame / inter-frame determination unit 4 directly uses the input image data to perform the signal processing in the intra-frame mode, or uses the immediately preceding frame (hereinafter, referred to as the previous frame) to perform the signal processing in the inter-frame mode. This is a circuit that determines whether or not to perform the determination, and the determination result is provided to the motion compensation processing unit 5. The motion compensation processing unit 5 is a circuit that detects the image data (hereinafter referred to as a search block) having the highest correlation with the input image data from the previous frame when performing the processing in the inter-frame mode.

The difference processing unit 6 is a circuit that performs difference calculation between the input image data and the search block detected by the motion compensation processing unit 5 and outputs the result as a difference block. The conversion processing unit 7 inputs the image data from the intra-frame / inter-frame determination unit 4 in the intra-frame mode, and inputs the difference block calculated by the difference processing unit 6 in the inter-frame mode to calculate the discrete cosine. This is a circuit that reduces the amount of image information by a conversion method such as conversion (hereinafter referred to as DCT conversion). The quantization processing unit 8 is the conversion processing unit 7.
This is a quantizing circuit that approximates the image data converted in (1) to discrete values, and its output is given to the variable length coding processing unit 9 and the dequantization processing unit 11. The inverse quantization processing unit 11 is a circuit that performs inverse quantization to restore the image data quantized by the quantization processing unit 8. The inverse transform processing unit 12 is a circuit for performing an inverse DCT transform on the image data dequantized by the inverse quantization processing unit 11, and its output is stored in the frame memory 13.

The variable length coding processing unit 9 is a circuit for sequentially fetching the image data quantized by the quantization processing unit 8 and performing variable length coding of the image data. The network interface processing unit 10 is a circuit that performs control for connecting the image processing apparatus to the communication network and sends the image data encoded by the variable length encoding processing unit 9 to the other party.

A series of operations for compressing and transmitting moving image data in the conventional image processing apparatus having the image compression processing means configured as described above will be described. First, the network interface processing unit 10 controls the connection of the communication channel with the partner device via the communication network to make the communication possible. After that, the image is input to the intra-frame / inter-frame determination unit 4 frame by frame. At this time, it is assumed that one frame image is divided into a certain size to create image data and the signal is processed. Next, the intra-frame / inter-frame determination unit 4 identifies, for each input image data of one frame, whether the image data is the same as the image data in the previously input frame or the image data in which the image content is changed. It is determined whether to perform the processing in step 1 or the interframe mode.
When the input image data is in the intra-frame mode, the intra-frame / inter-frame determination unit 4 transfers the input image data to the conversion processing unit 7, and when the input image data is in the inter-frame mode,
The input image data is output to the motion compensation processing unit 5.

When the image data is input to the motion compensation processing unit 5, the image data including the image data at the same position and its periphery held in the frame memory 13 is taken out, and the image processing using, for example, the minimum absolute difference value is performed. Method, the image data (search block) having the highest correlation with the input image data is detected. Then, the displacement (motion vector) of the position between the search block and the input image data is obtained. Next, the difference processing unit 6
Then, a difference block is created by calculating the difference between the input image data and the search block detected by the motion compensation processing unit 5, and the data is output to the conversion processing unit 7.

In the conversion processing unit 7, the input image data is input from the intra-frame / inter-frame determination unit 4 in the intra-frame mode, and the difference block is input from the difference processing unit 6 in the inter-frame mode. In any case, the conversion processing unit 7 performs the information amount reduction processing on the input data by a conversion method such as DCT conversion, and supplies the data to the quantization processing unit 8. The quantization processing unit 8 converts the DCT-transformed image data into discrete approximate values. Further, the inverse quantization processing unit 11 and the inverse transformation processing unit 12 create data for detecting the search block in the motion compensation processing unit 5. That is, the quantization processing unit 8 performs inverse quantization processing and inverse conversion processing of the quantized macroblock, outputs the image data processed here in the intra-frame mode, and processes it here in the inter-frame mode. The addition operation of the image data and the search block is performed, and the addition data is given to the frame memory 13. The frame memory 13 stores the image data output from the inverse conversion processing unit 12 at the same position as the image data of the previous frame. By repeating such a series of operations for a plurality of image data, one frame of compressed code data is transmitted to the partner device via the communication network.

[0009]

However, in the conventional image processing apparatus, there is a limit to the improvement of the processing capacity for processing all the image data of the input frame, and a communication network with a narrow transmission rate is used. There is a problem that more frames cannot be transmitted. Therefore, it has been difficult to reproduce smooth image data on the partner device.

The present invention has been made in view of such conventional problems, and reduces the compressed image data to be transmitted, increases the number of transmittable image frames, and reduces the communication rate of the communication network at a low rate. However, it is an object to realize an image processing device capable of reproducing a smooth image.

[0011]

According to the present invention, image data is input frame by frame, the image data is compressed by intra-frame coding or inter-frame coding, and the compressed image data is quantized and variable-length coded. An image processing apparatus having image compression processing means for performing an encoding process and outputting, wherein a motion comparison unit for inputting image data of a current frame and a previous frame respectively and generating difference data of images, and a motion comparison unit When the difference data is input and the value is less than the predetermined threshold value, the image of the current frame is determined to be a still image, an identification signal indicating that transmission of the compressed data of the current frame is omitted is output, and the difference data is equal to or more than the predetermined threshold value. In this case, a still judging unit for judging that the image of the current frame is in motion and giving the current image data to the image compression processing means is provided.

[0012]

According to the present invention having such a feature, the motion comparing section inputs the image data of the current frame and the image data of the previous frame, respectively, and generates image difference data. The still judging section
When the difference data of the motion comparison unit is input, when the value is less than a predetermined threshold value, the image of the current frame is determined to be a still image, and an identification signal indicating that transmission of the compressed data of the current frame is omitted is output. When the difference data is equal to or larger than the predetermined threshold value, it is determined that the image of the current frame is in motion. When the still judging unit judges that the image has moved, the current image data is given to the image compression processing means. Further, when the identification signal is output from the still judging unit, the image data of the corresponding frame is not sent to the receiving side at all, and the image processing of the next frame is started. In this way, the image data is compressed by intra-frame coding or inter-frame coding, and the image data is quantized and variable-length coded and output. This increases the transmission rate of the image by the omitted frames.

[0013]

DESCRIPTION OF THE PREFERRED EMBODIMENTS An image processing apparatus according to an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing the overall configuration of the image processing apparatus according to the present embodiment. The same parts as those in the conventional example are designated by the same reference numerals, and the description of the configuration of the image compression processing means and the detailed description of the blocks will be omitted. .
In the figure, image input (input image data) is input to the motion comparison unit 1 in units of one frame. The motion comparison unit 1 inputs the image data of the current frame to be processed and compares the previous frame data held in the frame memory 13 with the previous image data at the same position. Then, the difference value between the two image data is calculated using the sum of absolute differences between the previous image data and the current image data. The still judging unit 2 receives the difference data calculated by the motion comparing unit 1, compares the data with a predetermined threshold value, and judges whether the current image data is larger or smaller than the previous image data. To do.

When the still judging unit 2 judges that there is motion, it outputs the current image data as it is to the intra-frame / inter-frame judging unit 4. If it is determined that the movement is extremely small, the next image data is input without processing the current image data. In this case, the still judging unit 2 converts the identification signal indicating that the current frame is the same as the previous frame, that is, the still image, into a minimum amount of code and directly outputs it to the network interface processing unit 10. Network interface processing unit 1
0 outputs a still identification signal to the receiving side. Then, the motion comparing section 1 and the still judging section 2 repeat the same operation.

Next, the intra-frame / inter-frame judging unit 4 takes in the current image data and the image data of the preceding frame when the still judging unit 2 judges that there is motion. And 1
For each input image data of the frame, it is identified whether it is the same as the image data in the previous frame or changed image data, and it is determined whether the processing is performed in the intra-frame mode or the inter-frame mode. When the input image data is in the intra-frame mode, the intra-frame / inter-frame determination unit 4 transfers the input image data to the conversion processing unit 7, and when the input image data is in the inter-frame mode, the input image data is input to the motion compensation processing unit 5. Is output.

When the image data is input to the motion compensation processing unit 5, the image data including the image data at the same position and its periphery held in the frame memory 13 is taken out, and the image processing method using the minimum sum of absolute differences is used. Then, the search block having the highest correlation with the input image data is detected. Then, the motion vector between the search block and the input image data is obtained.
Next, the difference processing unit 6 creates a difference block by calculating the difference between the input image data and the search block detected by the motion compensation processing unit 5, and outputs the data to the conversion processing unit 7.

The conversion processing unit 7 inputs the input image data from the intra-frame / inter-frame judging unit 4 in the intra-frame mode, and the difference processing unit 6 in the inter-frame mode.
Input more difference block. In any case, the conversion processing unit 7 performs the information amount reduction process on the input data by a conversion method such as DCT conversion, and the data is quantized by the quantization processing unit 8
Give to. The quantization processing unit 8 quantizes the DCT-transformed image data. Further, the inverse quantization processing unit 11 and the inverse transformation processing unit 12 create data for detecting the search block in the motion compensation processing unit 5. That is, the quantization processing unit 8 performs inverse quantization processing and inverse conversion processing of the quantized macroblock, outputs the image data processed here in the intra-frame mode, and processes it here in the inter-frame mode. The addition operation of the searched image data and the search block is performed,
The addition data is given to the frame memory 13.

The frame memory 13 stores the image data output from the inverse conversion processing unit 12 at the same position as the image data of the previous frame. Then, the frame memory 13 gives the image data accumulated for each frame to the motion comparison unit 1 and the motion compensation processing unit 5. By repeating such a series of operations for a plurality of image data, one frame of compressed code data is transmitted to the partner device via the communication network. In particular, when the still determination unit 2 determines that the image of the current frame is the same as the previous frame, the identification signal is transmitted to the image processing device on the receiving side via the network interface processing unit 10. In this case, immediately after the identification signal is output in the frame cycle in which the same image is generated on the transmitting side, the image processing of the next frame starts. Therefore, the transmission of image data is shortened by the number of frames determined to be still, and the frame transmission rate per unit time is increased.

[0019]

As described above, according to the present invention, by providing the motion comparing section and the still judging section, the image data to be image-compressed and transmitted can be reduced and the number of frames that can be transmitted can be increased. it can. Therefore, it is possible to realize an image processing device that reproduces a smooth image even in a low-rate communication network.

[Brief description of drawings]

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

FIG. 2 is a block diagram showing a configuration example of a conventional image processing apparatus.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 motion comparison unit 2 still determination unit 4 intra-frame / inter-frame determination unit 5 motion compensation processing unit 6 difference processing unit 7 conversion processing unit 8 quantization processing unit 9 variable length coding processing unit 10 network interface processing unit 11 dequantization Processing unit 12 Inverse conversion processing unit 13 Frame memory

Claims (1)

[Claims] 1. Inputting image data on a frame-by-frame basis, compressing the image data by intra-frame coding or inter-frame coding, outputting the compressed image data by performing a quantization process and a variable length coding process. In the image processing apparatus having the image compression processing means, the motion comparison unit that inputs the image data of the current frame and the image data of the previous frame and generates the difference data of the image, and the difference data of the motion comparison unit are input, and the value thereof is input. Is less than a predetermined threshold value, the image of the current frame is determined to be a still image, and an identification signal indicating that the transmission of compressed data of the current frame is omitted is output. If the difference data is equal to or more than a predetermined threshold value, the image of the current frame is output. And a still judging unit for judging that there is motion and giving the current image data to the image compression processing means.