Method and apparatus for video encoding/decoding based on orthogonal transform and vector quantization

Abstract

Provided are a method and an apparatus for video encoding/decoding based on an orthogonal transform and vector quantization. A video encoding method includes determining whether an input image is a residual image, if the input image is a residual image, performing first encoding on the input image through transform/quantization and performing second encoding on the input image through vector quantization, comparing data obtained through the first encoding and data obtained through the second encoding, and selecting an encoding type based on the result of the comparison, generating mode information indicating the selected encoding type, and outputting data obtained according to the selected encoding type

Description

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other aspects of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the attached drawings, in which:

FIG. 1 is a block diagram of a related art video encoder;

FIG. 2 is a block diagram of a video encoder according to an exemplary embodiment of the present invention;

FIGS. 3A and 3B are views for explaining a vector quantization mode according to an exemplary embodiment of the present invention;

FIG. 4 is a flowchart illustrating a video encoding method used in the video encoder of FIG. 2 according to an exemplary embodiment of the present invention;

FIG. 5 is a block diagram of a video decoder according to an exemplary embodiment of the present invention; and

FIG. 6 is a flowchart illustrating a video decoding method used in the video decoder of FIG. 5 according to an exemplary embodiment of the present invention.

Claims

1. A video encoding method comprising: determining whether an input image is a residual image;if the input image is the residual image, performing first encoding on the input image through transform/quantization and performing second encoding on the input image through vector quantization;comparing first data obtained through the first encoding and second data obtained through the second encoding; andselecting an encoding type based on a result of the comparison, generating mode information indicating the selected encoding type, and outputting third data obtained according to the selected encoding type.

2. The video encoding method of claim 1, wherein the comparison of the first and second data comprises comparing a first bitrate of the first data and a second bitrate of the second data.

3. The video encoding method of claim 1, further comprising performing entropy encoding on the first-encoded input image and the second-encoded input image before the comparison of the first and second data, wherein the comparison of the first and second data comprises comparing a first bitrate of the first-encoded input image on which the entropy encoding is performed, and a second bitrate of the second-encoded input image on which the entropy encoding is performed.

4. The video encoding method of claim 1, wherein the comparison of the first and second data comprises comparing a first rate-distortion cost of the first data and a second rate-distortion cost of the second data.

5. The video encoding method of claim 4, wherein the first and second rate-distortion costs are obtained by performing inverse transform/quantization on the first-encoded input image and inverse vector quantization on the second-encoded input image, respectively.

6. The video encoding method of claim 1, wherein if the input image is not the residual image, only the first encoding is performed on the input image.

7. The video encoding method of claim 1, further comprising performing entropy-encoding on the third data and the generated mode information.

8. The video encoding method of claim 7, wherein: if the selected encoding type is the transform/quantization, the third data comprises quantized coefficients, and zigzag scanning is performed on the third data before the entropy-encoding is performed on the third data; andif the selected encoding type is the vector quantization, the third data comprises index information of a codebook having an image pattern that is most similar to a pattern of the input image, and the entropy-encoding is performed on the index information.

9. The video encoding method of claim 1, wherein the transform comprises an orthogonal transform.

10. A video encoder comprising: a determination unit which determines whether an input image is a residual image;a first encoding unit which performs transform/quantization on the input image;a second encoding unit which performs vector quantization on the input image if the input image is a residual image;a comparison unit which compares first data obtained by the first encoding unit and second data obtained by the second encoding unit; anda mode selection unit which selects an encoding type based on a result of the comparison, generates mode information indicating the selected encoding type, and outputs third data obtained according to the selected encoding type.

11. The video encoder of claim 10, wherein the comparison unit compares a first bitrate of the first data and a second bitrate of the second data.

12. The video encoder of claim 10, further comprising an entropy encoding unit which performs entropy encoding on the first-encoded input image and the second-encoded input image before the comparison of the first and second data, wherein the comparison of the first and second data comprises comparing a first bitrate of the first-encoded input image on which the entropy encoding is performed, and a second bitrate of the second-encoded input image on which the entropy encoding is performed.

13. The video encoder of claim 10, wherein the comparison unit compares a first rate-distortion cost of the first data and a second rate-distortion cost of the second data.

14. The video encoder of claim 13, further comprising an inverse quantization/inverse transform unit which performs inverse transform/quantization, and an inverse vector quantization unit which performs inverse vector quantization, wherein the first and second rate-distortion costs are obtained by performing the inverse transform/quantization on the transform/quantized input image and the inverse vector quantization on the vector quantized input image, respectively.

15. The video encoder of claim 10, wherein if the input image is not the residual image, only the transformation and quantization are performed on the input image.

16. The video encoder of claim 10, further comprising an entropy encoding unit which performs entropy-encoding on the third data and the generated mode information.

17. The video encoder of claim 16, wherein: if the selected encoding type is the transform/quantization, the third data comprises quantized coefficients, and zigzag scanning is performed on the third data before the entropy-encoding in performed on the third data; andif the selected encoding type is the vector quantization, the third data comprises index information of a codebook having an image pattern that is most similar to a pattern of the input image, and the entropy-encoding is performed on the index information.

18. The video encoder of claim 10, wherein the transform comprises an orthogonal transform.

19. A video decoding method comprising: performing entropy-decoding on an input bitstream comprising video data, motion information, and mode information indicating an encoding type of an input image;performing first decoding on the video data through inverse quantization/inverse transform or second decoding on the video data through inverse vector quantization based on the mode information; andadding video data that is motion-compensated or intraprediction-decoded using the motion information to the decoded video data, thereby generating reconstructed video data.

20. The video decoding method of claim 19, wherein if the input image is a residual image, the mode information indicates the encoding type of the input image, and wherein the encoding type is selected by performing first encoding on the input image through transform/quantization, performing second encoding on the input image through vector quantization, and comparing first data obtained through the first encoding and second data obtained through the second encoding.

21. The video decoding method of claim 20, wherein the comparison of the data comprises comparing a first bitrate of the first data and a second bitrate of the second data.

22. The video decoding method of claim 20, wherein the comparison of the first and second data comprises comparing a first bitrate of the first-encoded input image on which entropy encoding is performed, and a second bitrate of the second-encoded input image on which entropy encoding is performed.

23. The video decoding method of claim 20, wherein the comparison of the first and second data comprises comparing a first rate-distortion cost of the first data and a second rate-distortion cost of the second data.

24. The video decoding method of claim 19, wherein the inverse transform comprises an inverse orthogonal transform.

25. A video decoder comprising: an entropy-decoding unit which performs entropy-decoding on an input bitstream comprising video data, motion information, and mode information indicating an encoding type of an input image;a first decoding unit which performs first decoding on the video data through inverse quantization/inverse transform based on the mode information;a second decoding unit which performs second decoding on the video data through inverse vector quantization based on the mode information; anda video reconstruction unit which adds video data that is motion-compensated or intraprediction-decoded using the motion information to the decoded video data, thereby generating reconstructed video data.

26. The video decoder of claim 25, wherein if the input image is a residual image, the mode information indicates the encoding type of the input image, and wherein the encoding type is selected by performing first encoding on the input image through transform/quantization, performing second encoding on the input image through vector quantization, and comparing first data obtained through the first encoding and second data obtained through the second encoding.

27. The video decoder of claim 26, wherein the comparison of the data is performed by comparing a first bitrate of the first data and a second bitrate of the second data.

28. The video decoder of claim 26, wherein the comparison of the first and second data comprises comparing a first bitrate of the first-encoded input image on which entropy encoding is performed, and a second bitrate of the second-encoded input image on which entropy encoding is performed.

29. The video decoder of claim 26, wherein the comparison of the first and second data is performed by comparing a first rate-distortion cost of the first data and a second rate-distortion cost of the second data.

30. The video decoder of claim 25, wherein the inverse transform comprises an inverse orthogonal transform.

31. A computer-readable recording medium having recorded thereon a program for performing a video encoding method comprising: determining whether an input image is a residual image;if the input image is the residual image, performing first encoding on the input image through transform/quantization and performing second encoding on the input image through vector quantization;comparing first data obtained through the first encoding and second data obtained through the second encoding; andselecting an encoding type based on a result of the comparison, generating mode information indicating the selected encoding type, and outputting third data obtained according to the selected encoding type.

32. A computer-readable recording medium having recorded thereon a program for performing a video decoding method comprising: performing entropy-decoding on an input bitstream comprising video data, motion information, and mode information indicating an encoding type of an input image;performing first decoding on the video data through inverse quantization/inverse transform or second decoding on the video data through inverse vector quantization based on the mode information; andadding video data that is motion-compensated or intraprediction-decoded using the motion information to the decoded video data, thereby generating reconstructed video data.