KR980013419A - Encoders and decoders in image processing systems - Google Patents

Abstract

The present invention relates to an encoder and a decoder of an image processing system. Conventionally, a compression ratio is determined according to a signal processing method. Therefore, the present invention is characterized in that means for downsampling 208 is added to the input of the encoder, and when the sum of the high frequency components of the coefficients by discrete cosine transforming the downsampled signal exceeds a certain threshold value, And means for up-sampling (216) is added to the output terminal of the decoder. In the present invention, the digital image is compressed by performing down-sampling at the time of encoding, and up-sampling is performed at the time of decoding to recover the original bit stream, thereby achieving a higher compression rate than that of the prior art and obtaining a high quality image at the same bit rate .

Description

Encoders and decoders in image processing systems

Since this is a trivial issue, I did not include the contents of the text.

The present invention relates to an image compression technique, and more particularly, to an image compression system for improving the compression efficiency while maintaining a basic standard of MPEG1 and MPEG2, which are image compression methods, in an image processing system, Decoder.

Generally, the data obtained by digitally sampling the original screen in the image processing process is too much data to be used as it is, so it is compressed and used.

At this time, it is important to improve the compression ratio while maintaining a similar resolution to the original screen.

Therefore, in the image compression, a technique of compressing data to a degree that does not make a slight loss of data by using the human visual characteristic to the maximum, and a technique of compressing the data itself without loss are used.

FIG. 2 is a block diagram of a conventional encoder. As shown in FIG. 2, a discrete cosine transformer 101 for performing discrete cosine transform on the difference between a video input signal and a predicted signal and a quantizer 102 for quantizing an output signal of the discrete cosine transformer 101 A variable length encoder 103 for variable length coding the output signal of the quantizer 102 on the basis of a motion vector, a first-in first-out memory 103 for adjusting a quantization level of the variable length encoder 103, An inverse quantizer 105 for inversely quantizing an output signal of the quantizer 102, an inverse discrete cosine transformer 106 for inverse discrete cosine transforming the output signal of the inverse quantizer 105, , A motion vector is extracted from the video input signal, the motion vector is output to the variable length encoder 103, and the output signal of the inverse discrete cosine transformer 106 and the prediction signal are added together and the decoded video signal is input to the next frame And a motion estimation / compensation unit 107 for generating a prediction signal for the motion estimation / compensation unit 107.

FIG. 3 is a block diagram of a conventional decoder. As shown in FIG. 3, the buffer 111 includes a buffer 111 for temporarily storing input data, a variable length decoding unit 111 for performing variable length decoding on the output data of the buffer 111, An inverse quantizer 113 for inversely quantizing the output data of the variable length decoder 112 and an inverse discrete cosine transformer 111 for inversely discrete cosine transforming the output data of the inverse quantizer 113, And a motion compensator 115 for adjusting the data rate by checking the output of the inverse discrete cosine transformer 114. The inverse discrete cosine transformer 114 outputs the inverse discrete cosine transformer 114,

The operation of the conventional circuit will be described as follows.

First, when the luminance Y and the chrominance components Cb and Cr sampled as in the format of FIG. 1 are input, a difference between the luminance Y and the chrominance components Cb and Cr is obtained. The difference signal is input to the discrete cosine transformer 101, To the frequency domain.

Here, the discrete cosine transform (DCT) transforms image data in a space into an energy distribution on a frequency, and uses a property that general image data is concentrated in a low frequency region.

Thereafter, the quantizer 102 receives the discrete cosine transformed coefficients and performs a quantization process using the quantization table. The quantized coefficients are rearranged into a one-dimensional array so that data transmission is convenient, and a value other than "0" In order to concentrate the coefficients, the jig jig scan is performed and output.

Here, when the quantization process is performed, it is impossible to completely reproduce the original signal, and high-frequency components that are not sensitive to the eye are removed by using human visual characteristics.

Here, the quantization is a nonlinear operation by approximating a finite number of values of a discrete cosine transformed coefficient.

At this time, when the quantization process is performed, it is impossible to completely reproduce the original signal, and high-frequency components which are not sensitive to the eyes are removed using human visual characteristics.

After that, the variable length encoder 103 receives the motion vector MV of the motion estimation and compensation unit 107 and performs quantization of the quantized coefficients by performing variable length coding, which is a constant of entropy coding, to further compress .

In this case, variable length coding is a compression technique that allocates a long code to a symbol having a low frequency and allocates a short code to a symbol having a high frequency to reduce the data amount as a whole.

Thus, the first-in-first-out memory 104 temporarily stores the output data of the variable-length coder 103 and transmits the same.

At this time, an overflow of the first-in-first-out memory 104 is prevented by adjusting the quantization level of the quantizer 102 in accordance with the difference in the amount of input / output data in the first-in first-out memory 104.

Meanwhile, in compressing and transmitting the video data, the motion estimation and compensation unit 107 extracts a motion vector MV using the input video signal as an input, and at the same time, decodes the decoded data from the output signal of the inverse quantizer 106 And generates prediction data for the next frame as an input.

Accordingly, the motion vector MV generated by the motion estimation and compensation unit 107 generates the data to be decoded in the output signal from the inverse discrete cosine transformer 106 and the difference component input to the variable length encoder 101 .

The data having been subjected to the encoding process as described above is decoded in the opposite process in the decoder and displayed as an image on the screen of the display unit.

That is, when the coded data is input through the buffer 111, the variable length decoder 112 performs a decoding process to expand the compressed bitstream.

In this case, the motion vector MV and the quantization level are extracted and input to the inverse quantizer 113 and the motion compensation unit 115, respectively.

Accordingly, when the inverse quantizer 113 dequantizes and restores the output of the variable length decoder 112, the inverse discrete cosine transform unit 114 performs an inverse discrete cosine transform process to decode the original video data , The decoded data is displayed as an image on the screen through a display unit (not shown).

On the other hand, the motion compensation unit 115 checks the output of the inverse quantizer 114 based on the motion vector MV extracted by the variable length decoder 112, thereby preventing data overflow.

In the above process, the basic unit of data processing is 8x8 block unit.

However, MPEG1 and MPEG2, which are the standards of the current video compression technology, specify the specifications for the decoder block, but the specifications for the encoder block and the display end are not specified.

At this time, the image quality can be made similar to the existing standard, while increasing the compression ratio depending on how the encoder first processes the bit stream or how the final output bit stream is processed at the display end.

Therefore, conventionally, since the compression ratio is determined according to the signal processing method, there is a problem that the image quality is lowered in case of a complicated image.

In order to solve the conventional problems, the present invention provides an encoder and a decoder of an image processing system designed to improve compression rate by performing down-sampling in an encoding process to reduce data and performing interpolation in a decoding process to decode original data. And the like.

In order to achieve the above-mentioned object, the present invention is constituted by adding means for down-sampling to the input end of the encoder and means for up-sampling at the output end of the decoder.

In the present invention, the downsampling is performed to reduce the amount of data to improve the compression ratio and the upsampling is performed to decode the original data so that the resolution is substantially similar to that of the original image.

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

5 is a block diagram of an encoder according to an embodiment of the present invention. As shown therein, a discrete cosine transformer 201, a quantizer 202, a variable length decoder 203, a first-in first-out memory 204, The quantizer 205, the inverse discrete cosine transformer 206, and the motion estimation and compensation unit 207 are constructed in the same manner as the conventional encoder of FIG. 2, downsampling the video input signal and inputting to the discrete cosine transformer 201 And a downsampling unit 208 for generating a downsampling signal.

6 is a block diagram of a decoder according to an embodiment of the present invention. As shown in FIG. 6, a buffer 211, a variable length decoder 212, an inverse quantizer 213, an inverse discrete cosine transformer 214, And an interpolator 216 configured to be the same as the conventional decoder of FIG. 3 and to upsample the output signal of the inverse discrete cosine transformer 214 as an input.

The operation and effect of the embodiment according to the present invention will be described as follows.

First, a signal sampled in accordance with the MPEG format in the original image is input to the downsampling unit 208 and is downsampled to an average value between adjacent pixels.

At this time, if the downsampling process is performed on the bitstream of the original screen as shown in FIG. 4 (A), it is reduced to 1/4 of that of the original bitstream as shown in FIG. 4 (B).

Then, the discrete cosine transformer 201 performs discrete cosine transform on the bitstream of the original picture that has not passed through the downsampling unit 208, adds the high frequency components among the resultant values, and compares the sum value with a certain threshold value.

At this time, the discrete cosine transformer 201 discrete cosine transforms the output bit stream of the downsampling unit 208 and inputs the discrete cosine transform to the quantizer 202 when the summed value is smaller than a certain threshold value.

If the summed value is greater than a certain threshold value, the output bitstream of the downsampling unit 208 is subjected to discrete cosine transform and input to the quantizer 202, and then the bitstream information of the original image is further supplied to the quantizer 202 .

In this case, the discrete cosine transform is performed in units of 8.times.8 pixels, which is one block, and four blocks form one macroblock.

Therefore, the macroblock whose sum of the high-frequency components does not exceed the threshold value is calculated in units of one block by a down-sampling process.

Thereafter, the quantizer 202 approximates the discrete cosine transformed coefficient to a finite number value through a quantization process and outputs it.

Accordingly, when the variable length encoder 203 receives the motion vector MV of the motion estimation and compensation unit 207 and compresses the quantized coefficients by variable length coding, the first-inward selection memory 204 temporarily stores and transmits .

At this time, the overflow of the first-in-first-out memory 204 is prevented by adjusting the quantization level of the quantizer 102 according to the difference in the amount of input / output data of the first-in first-

Meanwhile, in compressing and transmitting the video data, the motion estimation and compensation unit 207 extracts the motion vector (MV) using the input video signal as input, and outputs the decoded data from the output signal of the inverse quantizer 206 And generates prediction data for the next frame as an input.

Accordingly, the motion vector MV generated by the motion estimation and compensation unit 207 is output to the variable length encoder 203, and the prediction data for the next frame is input to the discrete cosine transformer 201, And is used to generate decoded data in the output signal from the cosine transformer 206. [

The encoded data is decoded by the decoder. When the encoded data is input through the buffer 211, the variable length encoder 212 performs a decoding process to extend the compressed bitstream.

In this case, the motion vector MV and the quantization level are extracted and input to the inverse quantizer 213 and the motion compensation unit 215, respectively.

Accordingly, when the inverse quantizer 213 dequantizes and restores the output of the variable length coder 212, the inverse discrete cosine transformer 214 performs an inverse discrete cosine transform process.

Thereafter, the interpolator 216 receives the output data of the inverse discrete cosine transformer 214 and performs upsampling to recover the original bitstream. When the high-frequency component of the discrete cosine transformed coefficient at the time of coding exceeds a certain threshold value, When a bitstream is transmitted, the high-frequency component portion such as a boundary portion of the object is compensated by the bitstream.

Therefore, the original video data restored in the interpolating unit 216 is input into the display unit (not shown) and displayed as an image on the screen.

On the other hand, the motion compensation unit 215 checks the output of the inverse quantizer 214 based on the motion vector MV extracted by the variable length decoder 212, thereby preventing data overflow.

As described in detail above, according to the present invention, a digital image is compressed by downsampling during encoding, upsampling is performed during decoding to recover an original bitstream, thereby achieving a higher compression ratio than the conventional one, An image can be obtained.

Particularly, the present invention is suitable for a simple image such as a video communication system.

FIG. 1 is an illustration showing an image format; FIG.

Figure 2 is a block diagram of a conventional encoder.

3 is a block diagram of a conventional decoder; Fig.

FIG. 4 illustrates an example of downsampling in the present invention. FIG.

FIG. 5 is a block diagram of an encoder of the present invention. FIG.

FIG. 6 is a block diagram of a decoder of the present invention. FIG.

DESCRIPTION OF THE REFERENCE NUMERALS

The motion estimation and compensation unit 208 includes a downsampling unit 211 and a downsampling unit 211. The motion estimation and compensation unit 211 performs a downsampling operation on the motion compensated motion vector. A variable length decoder 212, an inverse quantizer 214, an inverse discrete cosine transformer 215, a motion compensator 216,

Claims (3)

Quantizing means for quantizing an output signal of the discrete cosine transforming means into a finite number of values; a quantizing means for quantizing the output signal of the discrete cosine transforming means; A first-in-first-out memory for temporarily storing an output signal of the variable length coding means and outputting the output signal; and an inverse quantization circuit for outputting an output signal of the inverse quantization means An inverse discrete cosine transform means for inversely discrete cosine transforming the output signal of the inverse quantizing means, a motion vector extracting means for extracting a motion vector from the video input signal and outputting the motion vector to the variable length coding means, And the decoded video signal is added to the input video signal. And a motion estimation and compensation means for generating a prediction signal for a negative frame. 2. The apparatus according to claim 1, wherein the discrete cosine transform means sums the high frequency components of the discrete cosine coefficients and, when the value exceeds a certain threshold value, transfers the output data of the downsampling means to the quantization means, And a signal is transmitted to the encoder. A variable length decoding means for performing variable length decoding on the output data of the buffer means and for extracting a quantization level and a motion vector; an inverse quantization means for inversely quantizing the output data of the variable length decoding means; Inverse discrete cosine transform means for performing inverse discrete cosine transform on the output data of the inverse quantization means and an output of the inverse discrete cosine transform means on the basis of the motion vector extracted by the variable length decoding means to adjust the data rate And a security means for upsampling an output signal of the inverse discrete cosine transform means and outputting an original video signal. ※ Note: It is disclosed by the contents of the first application.