KR950005624B1 - Image compressing & encoding apparatus and method - Google Patents

Abstract

The device increases an efficiency of a compression and eliminates noises by using a critical condition comparator. It comprises : (i) an input processor(22) which transforms analog signals of a image into digital signals; (ii) a block organizer(24) which organizes the output from the input processor(22) into block data; (iii) a frame memory(42); (iv) a critical condition comparator(44) which subtracts a previous frame block data of the frame memory(42) from the data of the block organizer(24), reduces data of a predicted error, and eliminates noises; (v) a transforming encoder(30); (vi) a quantizer(32); (vii) a reverse quantizer(36); (viii) a reverse transforming encoder(38); (ix) an adder(40).

Description

Image compression encoding apparatus and method

1 is a block diagram of a conventional video compression encoding apparatus.

2 is a block diagram of a video compression encoding apparatus according to the present invention.

3 is a concrete block diagram of the threshold condition comparator 44 in FIG.

4 is a flowchart for performing a subtraction and critical condition function according to the present invention.

5 is a shape diagram of a changing frame of a moving picture.

* Explanation of symbols for main parts of the drawings

22: input processor 24: block constructor

25: block configurator 30: conversion encoder

32: quantizer 36: inverse quantizer

38: inverse transform encoder 40: adder

42: frame memory

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video compression encoding apparatus and method for an image processing system such as a video telephone or a videoconferencing system. More particularly, the present invention relates to a compression encoding apparatus and method for compressing and transmitting a video signal using a threshold condition comparator.

In order to transmit a normal video signal to a counterpart through a telephone network, it is necessary to digitally convert the video signal and compress it to a small amount of digital data. A device for digitally converting and transmitting a video signal as described above is called a compression encoder, and a conventional video signal compression encoder is shown in FIG. 1.

1 is a circuit diagram of a conventional compression encoding apparatus for a video signal, and is a circuit diagram of a known compression encoding apparatus.

When the analog image signal is input to the input line 20 of the conventional compression encoding apparatus configured as shown in FIG. 1, the analog image signal is converted into digital by the input processor 22, and the digitally converted signal is Filtered and input to the block configurator 24. The digital video signal input to the block configurator 24 configures the input video signal in units of blocks in a scanning order.

At this time, from the frame memory 42, the video signal value of one block of the same size at the same position spatially in the previous frame is read from the internal memory and inputted to the subtractor 28. The subtractor 28 subtracts the video signal of the block unit output from the block configurator 24 and the output data of the frame memory 42 to calculate a difference at a spatially identical position.

These predictive error values are transform-coded by the transform encoder 30 and then quantized by the quantizer 32 and transmitted to the line 34.

On the other hand, in order to encode the next frame, the inverse quantizer 36 and the inverse transform encoder 38 perform the inverse processes of the transform encoder 30 and the quantizer 32, and the adder 40 adds the prediction value and the prediction error value to the frame memory. Save at 42.

Therefore, only the data from which the image block data previously transmitted is subtracted from the current image block is transmitted to the output line 34, and the frame memory 42 is transmitted to the current line 34 from the previously transmitted image block data. The added data is added and newly recorded. The receiving side may go through the reverse process of the transmitting side.

The prediction error values output from the subtractor 28 are intraframe 2D transformed by the transform encoder 30. In this case, DCT (Discrete Cosine Transform) and DFT (Discrete Fourier Transfor) can be used.

When the input processor 22 converts an analog image into 8-bit digital and inputs 8-bit image data in the compression encoding apparatus as shown in FIG. 1, the input processor 22 is reproduced by the adder 40 so that the frame memory 42 ) Should be 8 bits.

The reason for limiting the input video signal as a digital data of 8 bits are assigned to the 8 samples of the CCIR Recommendation 601. The video signal 0-255 (luminance signal level of the luminance signal as a standard for digital coding recommendations to the bits: ²⁸ = 255 This is because it is limited to the range of) to maintain compatibility with the system for processing video signals.

Therefore, the data stored in the frame memory 42 of FIG. 1 is inputted through the input processor 22 and the block configurator 24. The image data of one frame is subtracted by the subtractor 28, the transform encoder 30, and the quantizer ( 30), the inverse quantizer 36, the inverse transform encoder 38, and the adder 40 are data restored. Finally, the data stored in the frame memory 42 is data of 0 to 255 levels corresponding to 8 bits. It can be seen that it must be a value.

In the conventional compression encoding apparatus as described above, in the case of a still image or a motionless image having the same background, the prediction error value output from the subtractor 28 becomes zero, and noise is applied to the input signal or the camera shakes finely. Although the input video signal is a still image or a motionless image with the same background, when the prediction error value has a predetermined value other than zero, the total amount of data to be transmitted is increased, thereby reducing the efficiency of the video compression encoding apparatus. Or image quality is reduced by transmitting a video signal containing noise.

Accordingly, an object of the present invention is to provide a video encoding compression apparatus for solving the above problems.

Another object of the present invention is to provide an image compression encoding apparatus and method for increasing image compression efficiency by reducing the amount of data transmitted using a threshold condition comparator in an image processing system.

Another object of the present invention is to provide an apparatus and method for compressing and encoding an image to remove noise of data transmitted using a threshold condition comparator in an image processing system.

In order to achieve the above object, the present invention provides an input processor 22 for digitally converting and outputting an analog image signal, and a block configured to output the output data of the input processor 22 in block unit data according to a scanning order. A configurator 24, a frame memory 42 for outputting block data of the same size spatially the same position in the previous frame, and a transfer from the data of the block configurator 28 to the frame memory 42 A threshold condition comparator 44 which reduces the data amount of the prediction error output by subtracting the frame block data, removes the noise, and outputs the noise; and a transform encoder 30 which transforms and quantizes the output of the threshold condition comparator 44. And a quantizer 32, an inverse quantizer 36 and an inverse transform encoder 38 for inverse quantization and inverse transform encoding of the output of the quantizer 32, and the inverse transform encoding. It adds the output data of the frame memory 42 of 38 characterized in that consists of an adder 40 for outputting the reproduced data (RD).

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

2 is a block diagram of an image compression encoding apparatus according to the present invention. The input processor 22 digitally converts and outputs an analog image signal, and output data of the input processor 22 in block units in a scanning order. A block configurator 24 configured to output data, a frame memory 42 for outputting block data having the same size spatially the same position in all frames, and the frame memory from the data of the block configurator 28. Converts and encodes the output of the threshold condition comparator 44 and the threshold condition comparator 44 which reduces the data amount of the prediction error output by subtracting the previous frame block data output from 42 and removes the noise. Inverse quantizer 36 and inverse transform encoder 38 for inverse quantization and inverse transform encoding of the output of the quantizer 32 and the quantizer 32 and the output of the quantizer 32 And an adder 40 which adds the output of the inverse transform encoder 38 and the data of the frame memory 42 to output reproduction data RD.

FIG. 3 is a detailed block diagram of the threshold condition comparator 44 of FIG. 2. The input buffer 50 temporarily stores the image signal output in units of blocks from the block configurator 24 and the input buffer 50. Control unit 52 for reading out the signal stored in the control unit and performing a subtraction and threshold condition function to output a prediction error value, and an output buffer 54 for temporarily storing the prediction error value output from the control unit 52. .

4 is a flowchart for performing a subtraction and threshold condition comparison function according to the present invention.

5 is a diagram of changing frame shapes of moving images.

An embodiment of the present invention described above will be described in detail with reference to FIGS. 2, 3, 4, and 5.

When the analog video signal is input to the line 20 now, the input processor 22 converts the input analog video signal into 8-bit digital data, and filters the band. At this time, the block configurator 24 configures the input image signal in block units and outputs it to the threshold condition comparator 44.

The video compression encoding apparatus used in an image processing system such as a video telephone or a video conferencing system usually requires a DSP or a DSP for each block function of FIG. By assigning a control unit consisting of a CPU to process a signal and handing over the data processed by the control unit to a functional block of a next stage in a pipelined manner, the detailed operation will be described with reference to FIGS. In 24), the image signals in block units output in the scanning order are stored in the input buffer 50 of FIG.

In step (4a), the controller 52 reads the signal stored in the input buffer 50, and proceeds to step (4b), where one block of the same size is located at the same position spatially in the entire frame from the frame memory 42. Read video signal value of. In step (4c), the prediction error data is a prediction error value obtained by subtracting the block image data of the previous frame output from the frame memory 42 from the output of the block configurator 42, that is, the block of the current input image signal. The prediction error value X (i) is detected by the following equation (1).

X (i) = P (i, t)-P (i, t-1) (1)

Pi (i, t): the brightness of the pixel at the i th position of the block in the t th frame

P (i, t-1): brightness of the pixel at the i th position of the block in the t th frame

As shown in Equation (1), the size of the prediction error value X (i) becomes smaller when the background is the same in the two consecutive frames and is still or the image data is small in motion. In particular, in the case of image data corresponding to a still image or a part without motion in the background, the value of the predicted error value X (i) is 0 (zero), which is affected by effects such as noise added to the input image or minute shake of the camera. There is a case where 0 (zero) is not produced.

Therefore, in step (4d), the controller 52 determines that the largest value max [| X (i) | i = 1 to N] among the prediction error values for all the pixels in the block arranged in the block configurator 24 is determined. If it is smaller than the arbitrarily set threshold Tb or larger than the check-type Tb, the process proceeds to step 4e and the predicted error value detected in step 4c is stored in the output buffer 54 as it is. However, if the largest value max [X (i)] among the prediction errors for all pixels of the corresponding block unit is smaller than the arbitrarily set threshold Tb in step (4d), the process proceeds to step (4f). If the average value is smaller than the arbitrarily set threshold value Ta by adding the absolute values of all prediction error values X (i) in the block arranged in units of blocks in step (24), proceed to step (4g). The relevant block is regarded as a stationary background image or a motionless image data, and the pixel value of the entire block is replaced with the pixel value of the corresponding block of the previous frame, resulting in a prediction error value X (i) ( The values of i = 1 to N) are all zeroed and passed to the conversion encoder 30 through the output buffer 54. In this way, the total amount of data to be coded and transmitted is further reduced, and noise is removed in units of blocks.

However, if the mean value is greater than the arbitrarily set threshold value Ta by adding the absolute values of all prediction error values X (i) in the block in step (4f), the process proceeds to step (4e) and the prediction detected in step (4c). The error value is stored in the output buffer 54 and passed to the conversion encoder 34.

Here, the product of comparing the threshold values Ta and Tb in steps (4d) and (4f) is as follows. That is, the t-1 th frame in FIG. 5A is changed to the t th frame as in FIG. 5B, and the current block of the t th frame has a slight movement of the object from the previous frame. When only some pixel values in the block are changed, the average value of the sum of absolute values of all prediction error values X (i) is small for all the pixels in the block as shown in Equation (2). It may come out and become smaller than the threshold Ta.

In this case, if the coding is performed after replacing the predicted error value X (i) value for all pixels in the block with 0 (zero) and transmitting it, the receiver will reverse the display and display the movement of the object. Despite this, the block is restored to the background image data of the previous frame with no movement. Therefore, if you look at the entire frame, only the block is shown as the background image instead of the moving object. Prevent the occurrence of phenomena.

In the step (4b), if the value of max [X | This means that the data has changed significantly from the image data value. In addition, the noise is spread throughout the block Can be used to filter out cases where the error is greater than Ta, which makes it possible to misinterpret the threshold comparison.

The threshold values Ta and Tb can be determined experimentally. After setting the values Ta and Tb to an arbitrary value, the receiver is not affected by noise by subjective evaluation based on the restored screen. It is determined within the limit that does not occur, and it can be determined experimentally as Ta = 3 and Tb = 16.

The prediction error values thus detected are transform-coded by the transform encoder 30 and then quantized by the quantizer 32 and transmitted to the line 34.

On the other hand, in order to encode the next frame, the inverse quantizer 36 and the inverse transform encoder 38 perform a reverse process of the transform encoder 30 and the quantizer 32, and the adder 40 adds the prediction value and the prediction error value to the frame memory. Save at 42.

Therefore, only the data from which the image block data previously transmitted is subtracted from the current image block is transmitted to the output line 34, and the frame memory 42 is transmitted to the current line 34 from the previously transmitted image block data. The added data is added and newly recorded. The receiving side may go through the reverse process of the transmitting side.

As described above, when the video signal is compressed and transmitted in an image processing system such as a video telephone or a video conference system, the prediction error value of the current frame is detected. When there is no background or motion, the prediction error value is zero and the background is different. If there is a change encoding by using the prediction error value of the current frame and the previous frame as it is, there is an advantage that not only increases the compression efficiency by reducing the total amount of data to be transmitted but also removes the noise in units of blocks.

Claims (3)

A video compression encoding apparatus of an image processing system, comprising: an input processor (22) for digitally converting an analog image signal and outputting the same; and output data of the input processor (22) according to a scanning order in block unit data A block configurator 24, a frame memory 42 for outputting block data of the same size spatially the same in all frames, and data from the block configurator 28 to the frame memory 42 A threshold condition comparator 44 for reducing the amount of data of the predicted error output by subtracting the previous frame block data, and removing the noise for output, and a transform encoder for transform-coding and quantizing the output of the threshold condition comparator 44. 30 and the quantizer 32, an inverse quantizer 36 and an inverse transform encoder 38 for inverse quantization and inverse transform encoding of the output of the quantizer 32, and the inverse An adder 40 which adds the output of the encoding encoder 38 and the data of the frame memory 42 to output the reproduction data RD, and temporarily stores the signal subtracted and output by the subtractor 28 in units of blocks. A control unit 52 which reads an input buffer 50, a signal stored in the input buffer 50, performs a subtraction and a threshold condition function, and outputs a prediction error value, and a prediction error value output from the control unit 52. And an output buffer (54) for temporarily storing, and a frame memory (56) connected to the controller (52) for storing data of two frames of a current frame and a previous frame. The method of claim 1, further comprising: an input buffer 50 for temporarily storing image signals output in units of blocks from the block configurator 24, and a subtraction and threshold condition function by reading signals stored in the input buffer 50; And an output buffer (54) for temporarily storing the predicted error value output from the controller (52). An image compression encoding method of an image processing system, comprising: a prediction error value detection step of detecting a prediction error value by subtracting a video signal of a current frame and a previous frame arranged in block units; and detecting the prediction error value in a detection step Prediction error characterized in that it comprises a prediction error value determination process of receiving a prediction error value and outputting the prediction error value as a zero value when there is no background or motion, and outputting the detected prediction error value as it is when the background is different and there is motion. Value detection method.