Prediction image generating method and apparatus using single coding mode for all color components, and image and video encoding/decoding method and apparatus using the same

• US 20060203913A1
• Filed: 02/28/2006
• Published: 09/14/2006
• Est. Priority Date: 02/28/2005
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Birinoy, Dmitri, Kim, Hyun Mun, Cho, Daesung, Kim, Wooshik
• Granted Patent: US 8,019,001 B2
• Time in Patent Office: Days
• US Class Current: 375/240.160
• CPC Class Codes: H04N 19/103 Selection of coding mode or of prediction mode
• CPC Class Codes: H04N 19/107 between spatial and temporal predictive coding, e.g. picture refresh
• CPC Class Codes: H04N 19/147 according to rate distortion criteria rate-distortion as a criterion for motion estimation H04N19/567
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/186 the unit being a colour or a chrominance component
• CPC Class Codes: H04N 19/51 Motion estimation or motion compensation
• CPC Class Codes: H04N 19/593 involving spatial prediction techniques

Abstract

A prediction image generation method and apparatus using a single coding mode for all color components and image and video encoding and decoding methods and apparatuses using the same are provided. The prediction image generation apparatus includes: a spatial prediction image generation unit applying an identical prediction direction for each image component in an original image composed of at least two image components, and obtaining a spatial prediction image from a pixel spatially close to a pixel block of a predetermined size of a current frame; a temporal prediction image generation unit applying an identical block size, motion vector, and motion interpolation method for each image component in an original image composed of at least two image components, and obtaining a temporal prediction image by estimating motion in units of blocks of a predetermined size between a previous frame and a current frame of each of the image components; an encoding mode selection unit selecting an encoding mode by using the spatial prediction image and the temporal prediction image; and a single mode prediction image generation unit generating a prediction image by identically applying the encoding mode selected in the encoding mode selection unit to each component. By increasing the redundancy of information between prediction error signals of color components, the encoding efficiency can be enhanced in the method and apparatus.

Claims

1. A spatial prediction image generation method comprising:

applying an identical prediction direction for each image component in an original image composed of at least two image components; and

obtaining a spatial prediction image from a pixel spatially close to a pixel block of a predetermined size of a current frame.

2. The method of claim 1, wherein the size of the pixel block is 4×

4, and any one of 9 prediction directions is determined to be the prediction direction.

26. A computer readable recording medium having embodied thereon a computer program for executing the method of claim 1.

3. A temporal prediction image generation method comprising:

applying an identical block size, motion vector, and motion interpolation method for each image component in an original image composed of at least two image components; and

obtaining a temporal prediction image by estimating motion in units of blocks of a predetermined size between a previous frame and a current frame of each of the image components.

27. A computer readable recording medium having embodied thereon a computer program for executing the method of claim 3.

4. A prediction image generation method comprising:

applying an identical prediction direction for each image component in an original image composed of at least two image components;

obtaining a spatial prediction image from a pixel spatially close to a pixel block of a predetermined size of a current frame;

applying an identical block size, motion vector, and motion interpolation method for each image component in an original image composed of at least two image components;

obtaining a temporal prediction image by estimating motion in units of blocks of a predetermined size between a previous frame and a current frame of each of the image components;

selecting an encoding mode using the spatial prediction image and the temporal prediction image; and

generating a prediction image by identically applying the selected encoding mode to each component.

5. The method of claim 4, wherein, in the selection of the encoding mode, the bit quantity and/or distortion of the picture quality of each of the spatial prediction image and the temporal prediction image are calculated and an encoding mode corresponding to the smallest sum obtained by adding the bit quantity and/or distortion of the picture quality for each of the spatial prediction image and the temporal prediction image is selected as the encoding mode.

28. A computer readable recording medium having embodied thereon a computer program for executing the method of claim 4.

6. A prediction image generation apparatus comprising:

a spatial prediction image generation unit applying an identical prediction direction for each image component in an original image composed of at least two image components, and obtaining a spatial prediction image from a pixel spatially close to a pixel block of a predetermined size of a current frame;

a temporal prediction image generation unit applying an identical block size, motion vector, and motion interpolation method for each image component in an original image composed of at least two image components, and obtaining a temporal prediction image by estimating motion in units of blocks of a predetermined size between a previous frame and a current frame of each of the image components;

an encoding mode selection unit selecting an encoding mode by using the spatial prediction image and the temporal prediction image; and

a single mode prediction image generation unit generating a prediction image by identically applying the encoding mode selected in the encoding mode selection unit to each component.

7. The apparatus of claim 6, wherein the encoding mode selection unit comprises:

a spatial bit quantity and/or picture quality distortion calculation unit calculating the bit quantity and/or distortion of the picture quality of the spatial prediction image;

a temporal bit quantity/picture quality distortion calculation unit calculating the bit quantity and distortion of the picture quality of the temporal prediction image; and

a performance comparison unit comparing values obtained by adding the bit quantity and/or the picture quality distortion of each of the spatial prediction image and the temporal prediction image, and selecting an encoding mode corresponding to the smaller value.

8. An image encoding apparatus using a single coding mode for all color components comprising:

a prediction image generation unit generating a prediction image by using an identical encoding mode for each image component in an original image composed of at least two image components;

a residue generation unit generating a residue corresponding to the difference between the original image and the prediction image; and

an encoding unit generating a bitstream by encoding the residue data.

9. The apparatus of claim 8, wherein the prediction image generation unit comprises:

a spatial prediction image generation unit obtaining a prediction image by applying an identical prediction direction from a pixel spatially adjacent to a pixel block of a current frame for each image component of the original image when performing an intra method;

a temporal prediction image generation unit obtaining a prediction image by applying identical motion estimation motion interpolation methods in units of blocks of identical size between a previous frame and a current frame for each image component of the original image when performing an inter method;

an encoding mode selection unit selecting an encoding mode by using the spatial prediction image and the temporal prediction image; and

a single mode prediction image generation unit generating a prediction image by identically applying the encoding mode selected in the encoding mode selection unit to each component of the color image.

10. The apparatus of claim 9, wherein the encoding mode selection unit comprises:

a spatial bit quantity and/or picture quality distortion calculation unit calculating the bit quantity and/or distortion of the picture quality of the spatial prediction image;

a temporal bit quantity and/or picture quality distortion calculation unit calculating the bit quantity and/or distortion of the picture quality of the temporal prediction image; and

a performance comparison unit comparing values obtained by adding the bit quantity and/or the picture quality distortion of each of the spatial prediction image and the temporal prediction image, and selecting an encoding mode corresponding to the smaller value.

11. The apparatus of claim 8, further comprising a residue transform unit performing a residue transform on the residue generated by the residue generation unit using a correlation between residues of the image components,

wherein the encoding unit generates the residue of each image component as a bitstream.

12. The apparatus of claim 11, further comprising a transform/quantization unit performing transform and quantization on the residue transformed by the residue transform unit, in units of blocks of a predetermined size,

wherein the encoding unit generates each of the transformed and quantized coefficients as a bitstream.

13. An image encoding method using a single coding mode for all color components comprising:

generating a prediction image by using an identical encoding mode for each image component in an original image composed of at least two image components;

generating a residue corresponding to the difference between the original image and the prediction image; and

generating a bitstream by encoding the residue data.

14. The method of claim 13, wherein the generating of the prediction image comprises:

obtaining a prediction image by applying an identical prediction direction from a pixel spatially adjacent to a pixel block of a current frame for each image component of an original image composed of at least two image components;

obtaining a prediction image by applying an identical block size, motion vector, and motion interpolation method for each image component in an original image composed of at least two image components, and by estimating motion in units of blocks of a predetermined size between a previous frame and a current frame of each of the image components;

selecting an encoding mode by using the spatial prediction image and the temporal prediction image; and

generating a prediction image by identically applying the selected encoding mode to each component of the color image.

15. The method of claim 14, wherein in the selecting of the encoding mode, the bit quantity and/or distortion of the picture quality of each of the spatial prediction image and the temporal prediction image are calculated and an encoding mode corresponding to a smallest sum of the bit quantity and/or distortion of the picture quality of each of the spatial prediction image and the temporal prediction image is selected as an encoding mode.

16. The method of claim 13, further comprising performing a residue transform on the generated residue using a correlation between residues of the image components,

wherein, in the generating of the bitstream, the transformed residue for each image component is generated as a bitstream.

17. The method of claim 16, further comprising performing a transform and quantization on the transformed residue in units of blocks of a predetermined size,

wherein, in the generating of the bitstream, each of the transformed and quantized coefficients is generated as a bitstream.

29. A computer readable recording medium having embodied thereon a computer program for executing the method of claim 13.

18. An image decoding apparatus using a single coding mode for all color components comprising:

a data restoration unit restoring residue information by decoding the residue information from a bitstream when the difference between an original image composed of at least two image components and a prediction image in the original image is a residue; and

a prediction compensation unit restoring an image by adding a prediction image generated by applying an identical encoding mode to the restored residue data.

19. The apparatus of claim 18, wherein the prediction compensation unit comprises:

a spatial prediction compensation unit restoring an image by adding to the restored residue data a prediction image obtained by applying an identical prediction direction from a pixel spatially close to a pixel block of a current frame for each image component of the original image, when performing an intra method; and

a temporal prediction compensation unit restoring an image by adding to the restored residue data a prediction image obtained by applying identical motion estimation motion and interpolation methods in units of blocks of identical size between a previous frame and a current frame for each image component of the original image, when performing an inter method.

20. The apparatus of claim 18, further comprising an inverse residue transform unit performing inverse transform on residue transformed data using a correlation between residues of the image components if the data restored in the data restoration unit is residue transformed data,

wherein the prediction compensation unit performs prediction compensation by adding to the inverse residue transformed residue a prediction image generated using an identical encoding mode for each image component.

21. The apparatus of claim 18, further comprising an inverse quantization/inverse transform unit performing inverse quantization and an inverse transform on transformed and quantized data in units of blocks of a predetermined size if the data restored in the data restoration unit is transformed and quantized residue data,

wherein the prediction compensation unit performs prediction compensation by adding to the inverse quantized and inverse transformed residue a prediction image generated using an identical encoding mode, for each image component.

22. An image decoding method using a single coding mode for all color components comprising:

restoring residue information by decoding the residue information from a bitstream when the difference between an original image composed of at least two image components and a prediction image in the original image is a residue; and

restoring an image by adding a prediction image generated by applying an identical encoding mode to the restored residue data.

23. The method of claim 22, wherein the restoring of the image comprises:

obtaining a prediction image by applying an identical prediction direction from a pixel spatially close to a pixel block of a current frame for each image component of the original image, when performing an intra method; and

obtaining a prediction image by applying identical motion estimation and motion interpolation methods in units of blocks of identical size between a previous frame and a current frame for each image component of the original image, when performing an inter method.

24. The method of claim 22, further comprising performing an inverse transform on the generated residue using a correlation between residues of the image components,

wherein an image is restored by adding to the inverse residue transformed residue a prediction image generated using an identical encoding mode for each image component.

25. The method of claim 22, further comprising performing inverse quantization and an inverse transform on the generated residue in units of blocks of a predetermined size,

wherein an image is restored by adding to the inverse quantized and inverse transformed residue a prediction image generated using an identical encoding mode, for each image component.

30. A computer readable recording medium having embodied thereon a computer program for executing the method of claim 22.

31. A prediction image generation method for an image having at least two image components, comprising:

obtaining a first prediction image for the at least two image components using a first image prediction method;

obtaining a second prediction image for at least two image components using a second image prediction method;

selecting an encoding mode using the first and the second prediction images; and

generating a prediction image by applying the selected encoding mode to each component.

32. The method of claim 31, wherein one of the first and the second image prediction methods is a spatial prediction image method.

33. The method of claim 31, wherein one of the first and the second image prediction methods is a temporal prediction image method.

34. The method of claim 31, wherein, in the selection of the encoding mode, a bit quantity and/or distortion of the picture quality of each of the first prediction image and the second prediction image are calculated and an encoding mode corresponding to the smallest sum obtained by adding the bit quantity and/or distortion of the picture quality for each of the first prediction image and the second prediction image is selected as the encoding mode.

35. A prediction image generation apparatus comprising:

an image generation unit to generate a first and a second prediction images for at least two image components;

an encoding mode selection unit selecting an encoding mode by using the first and the second prediction images; and

a single prediction image generation unit generating a prediction image by using the encoding mode selected in the encoding mode selection unit to each component.

36. The apparatus for claim 35, wherein the image generation unit comprises;

a spatial prediction image part; and

a temporal prediction image generation part.

37. The method of claim 35, wherein the encoding mode selection unit to calculate a bit quantity and/or distortion of the picture quality of each of the first prediction image and the second prediction image and an encoding mode corresponding and select the encoding mode having the smallest sum obtained by adding the bit quantity and/or distortion of the picture quality for each of the first prediction image and the second prediction image.