Methods and apparatus for illumination compensation of intra-predicted video

1. An apparatus, comprising:

a hardware encoder for encoding a picture using non-local intra prediction, wherein non-local intra prediction uses decoded areas of a current picture as predicted data for current data being encoded, and wherein said encoder includes an implicit illumination compensation processing portion for applying illumination compensation to predicted data obtained from the non-local intra prediction,

wherein the illumination compensation is modeled to obtain a resultant model, and parameters of the resultant model are derived using linear regression from a non-local picture data prediction technique that minimizes a distance between luminance and chrominance data to be encoded and luminance and chrominance data of a weighted and offset predicted signal and uses information located beyond local neighboring data within a decoded portion of the picture, wherein the parameters of the resultant model are implicitly signaled to a corresponding decoder.

2. The apparatus of claim 1 , wherein the illumination compensation is applied adaptively to the predicted data, the predicted data having been predicted by at least one of a displaced intra prediction and a template matching prediction.

3. The apparatus of claim 1 , wherein the resultant model is based on at least one of a polynomial function, a linear function, and a nonlinear function.

4. The apparatus of claim 3 , wherein the resultant model is a linear model based on a first order polynomial.

5. The apparatus of claim 4 , wherein the first order polynomial is an offset only function based on an additive constant.

6. The apparatus of claim 3 , wherein parameters of the illumination compensation are solved using a least square function.

7. The apparatus of claim 3 , wherein the non-local intra prediction is applied to at least one of luminance data, color data and depth view data.

8. A method, comprising:

encoding a picture using non-local intra prediction, wherein non-local intra prediction uses decoded areas of a current picture as predicted data for current data being encoded, and wherein said encoding step includes applying illumination compensation to predicted data obtained from the non-local intra prediction,

wherein the illumination compensation is modeled to obtain a resultant model, and parameters of the resultant model are derived using linear regression from a non-local picture data prediction technique that minimizes a distance between luminance and chrominance data to be encoded and luminance and chrominance data of a weighted and offset predicted signal and uses information located beyond local neighboring data within a decoded portion of the picture, wherein the parameters of the resultant model are implicitly signaled to a corresponding decoder.

9. The method of claim 8 , wherein the illumination compensation is applied adaptively to the predicted data, the predicted data having been predicted by at least one of: a displaced intra prediction and a template matching prediction.

10. The method of claim 8 , wherein the resultant model is based on at least one of: a polynomial function, a linear function, and a nonlinear function.

11. The method of claim 10 , wherein the resultant model is a linear model based on a first order polynomial.

12. The method of claim 11 , wherein the first order polynomial is an offset only function based on an additive constant.

13. The method of claim 10 , wherein parameters of the illumination compensation are solved using a least square function.

14. The method of claim 10 , wherein the non-local intra prediction is applied to at least one of luminance data, color data and depth view data.

15. An apparatus, comprising:

a hardware decoder for decoding a picture using non-local intra prediction, wherein non-local intra prediction uses decoded areas of a current picture as predicted data for current data being decoded, and wherein said decoder includes an implicit illumination compensation processing portion for applying illumination compensation to predicted data obtained from the non-local intra prediction,

wherein the illumination compensation is modeled to obtain a resultant model, and parameters of the resultant model are derived using linear regression from a non-local picture data prediction technique that minimizes a distance between luminance and chrominance data to be encoded and luminance and chrominance data of a weighted and offset predicted signal and uses information located beyond local neighboring data within a decoded portion of the picture, wherein the parameters of the resultant model are implicitly determined.

16. The apparatus of claim 15 , wherein the illumination compensation is applied adaptively to the predicted data, the predicted data having been predicted by at least one of: a displaced intra prediction and a template matching prediction.

17. The apparatus of claim 15 , wherein the resultant model is based on at least one of: a polynomial function, a linear function, and a nonlinear function.

18. The apparatus of claim 17 , wherein the resultant model is a linear model based on a first order polynomial.

19. The apparatus of claim 18 , wherein the first order polynomial is an offset only function based on an additive constant.

20. The apparatus of claim 17 , wherein parameters of the illumination compensation are solved using a least square function.

21. The apparatus of claim 17 , wherein the non-local intra prediction is applied to at least one of: luminance data, color data and depth view data.

22. A method, comprising:

decoding a picture using non-local intra prediction, wherein non-local intra prediction uses decoded areas of a current picture as predicted data for current data being decoded, and wherein said decoding step includes applying illumination compensation to predicted data obtained from the non-local intra prediction,

wherein the illumination compensation is modeled to obtain a resultant model, and parameters of the resultant model are derived using linear regression from a non-local picture data prediction technique that minimizes a distance between luminance and chrominance data to be encoded and luminance and chrominance data of a weighted and offset predicted signal and uses information located beyond local neighboring data within a decoded portion of the picture, wherein the parameters of the resultant model are implicitly determined.

23. The method of claim 22 , wherein the illumination compensation is applied adaptively to the predicted data, the predicted data having been predicted by at least one of: a displaced intra prediction and a template matching prediction.

24. The method of claim 22 , wherein the resultant model is based on at least one of: a polynomial function, a linear function, and a nonlinear function.

25. The method of claim 24 , wherein the resultant model is a linear model based on a first order polynomial.

26. The method of claim 25 , wherein the first order polynomial is an offset only function based on an additive constant.

27. The method of claim 24 , wherein parameters of the illumination compensation are solved using a least square function.

28. The method of claim 24 , wherein the non-local intra prediction is applied to at least one of: luminance data, color data and depth view data.