Search

EP-4736454-A1 - REFERENCE SAMPLE INTERPOLATION FOR INTRA PREDICTION

EP4736454A1EP 4736454 A1EP4736454 A1EP 4736454A1EP-4736454-A1

Abstract

A method and an apparatus for encoding or decoding a video are provided. For at least one pixel of a block of a picture, a predictor sample is determined based on an intra prediction along a non-diagonal direction, and at least four reference samples of the predictor sample are obtained from among a plurality of decoded pixels neighboring the block. A value for the predictor sample is obtained based on an interpolation that uses the at least four reference samples, and the block is encoded or decoded based on the value. In some embodiments, the interpolation is based on a linear interpolation and an update term that can be computed on the fly.

Inventors

  • RATH, Gagan Bihari
  • LE LEANNEC, FABRICE
  • Urban, Fabrice
  • RACAPE, FABIEN

Assignees

  • InterDigital CE Patent Holdings, SAS

Dates

Publication Date
20260506
Application Date
20240610

Claims (20)

  1. CLAIMS 1. A method, comprising: determining, for at least one pixel of a block of a picture, a predictor sample based on an intra prediction along a non-diagonal direction, obtaining at least four reference samples of the predictor sample from among a plurality of decoded pixels neighboring the block, obtaining a value for the predictor sample based on an interpolation that uses the at least four reference samples, wherein the interpolation is based on a filtering implemented as a linear interpolation and an update term. decoding the block based on the value.
  2. 2. An apparatus, comprising one or more processors, wherein said one or more processors is operable to determine, for at least one pixel of a block of a picture, a predictor sample based on an intra prediction along a non-diagonal direction, obtain at least four reference samples of the predictor sample from among a plurality of decoded pixels neighboring the block, obtain a value for the predictor sample based on an interpolation that uses the at least four reference samples, wherein the interpolation is based on a filtering implemented as a linear interpolation and an update term. decode the block based on the value.
  3. 3. A method, comprising: determining, for at least one pixel of a block of a picture, a predictor sample based on an intra prediction along a non-diagonal direction, obtaining at least four reference samples of the predictor sample from among a plurality of decoded pixels neighboring the block, obtaining a value for the predictor sample based on an interpolation that uses the at least four reference samples, wherein the interpolation is based on a filtering implemented as a linear interpolation and an update term. encoding the block based on the value.
  4. 4. An apparatus, comprising one or more processors, wherein said one or more processors is operable to determine, for at least one pixel of a block of a picture, a predictor sample based on an intra prediction along a non-diagonal direction, obtain at least four reference samples of the predictor sample from among a plurality of decoded pixels neighboring the block, obtain a value for the predictor sample based on an interpolation that uses the at least four reference samples, wherein the interpolation is based on a filtering implemented as a linear interpolation and an update term. encode the block based on the value.
  5. 5. The method of claim 1 or 3 or the apparatus of claim 2 or 4, wherein the interpolation is based on a 4-tap filtering or 6-tap filtering.
  6. 6. The method of claim 5, further comprising determining, from among the at least four reference samples, two reference samples nearest to a location of the predictor sample, one on each side of the location of the predictor sample, and wherein the linear interpolation uses the two nearest reference samples.
  7. 7. The method of claim 6, further comprising: determining filtered values for the two nearest reference samples, determining an absolute value of a difference between the filtered values of the two nearest reference samples, comparing the absolute value with a given value.
  8. 8. The method of claim 7, wherein responsive to a determination that the absolute value is below than or equals the given value, the at least one update term is set to 0.
  9. 9. The method of claim 1, 3 or 5-8, wherein if a coding mode of the block is among a given set of coding modes, the interpolation is based on a filtering wherein coefficients of the filter are calculated based on a fractional part of an horizontal offset of a position of the predictor sample from the pixel else the interpolation is based on a 6-tap cubic filter or a smoothing filter whose coefficients are pre- computed and stored in table.
  10. 10. The method of claim 9, wherein the given set comprises at least one of the ISP, MRL, TIMD, DIMD, SGPM, or wide angle.
  11. 11. The method of any of one claims 5-10, whose coefficients of the 4-tap or 6-tap filter are determined once per row of the block.
  12. 12. The method of any one of claims 5-11, wherein coefficients of the 4-tap or 6-tap filter are determined based on a fractional part of an horizontal offset of a position of the predictor sample from the pixel.
  13. 13. The method of claim 1, 3 or 5-12, wherein the interpolation is based on a linear interpolation and at least one update term.
  14. 14. The method of claim 13, wherein the at least one update term is determined using the at least four reference samples and a factor that depends on a fractional part of an horizontal offset of a position of the predictor sample from the pixel.
  15. 15. The method or the apparatus of claim 14, wherein the factor is determined once per row of the block.
  16. 16. The method of any of claims 5-13, wherein the linear interpolation uses the filtered values of the two nearest reference samples.
  17. 17. The method of claim 16, wherein an interpolation flag indicates whether the filtered or unfiltered values of the two nearest reference samples are used for the linear interpolation.
  18. 18. The method of claim 17, wherein the interpolation flag further indicates whether the filtered or unfiltered values of the two nearest reference samples are used for the update term.
  19. 19. The method of claim 17 or 18, wherein the interpolation flag is enabled based on a size of the block and an angle of the non-diagonal direction of the intra prediction.
  20. 20. The method of one of claims 17-19, wherein responsive to a determination that the absolute value is below than or equals the given value and the interpolation flag is enabled, the interpolation is a linear interpolation.

Description

REFERENCE SAMPLE INTERPOLATION FOR INTRA PREDICTION This application claims the priority to European Application No. EP23306037.5, filed on 28 June 2023, which is incorporated herein by reference in its entirety. TECHNICAL FIELD The present embodiments generally relate to video compression. The present embodiments relate to a method and an apparatus for encoding or decoding an image or a video. More particularly, the present embodiments relate to improving intra coding. BACKGROUND To achieve high compression efficiency, image and video coding schemes usually employ prediction and transform to leverage spatial and temporal redundancy in the video content. Generally, intra or inter prediction is used to exploit the intra or inter picture correlation, then the differences between the original block and the predicted block, often denoted as prediction errors or prediction residuals, are transformed, quantized, and entropy coded. In inter prediction, motion vectors used in motion compensation are often predicted from motion vector predictor. To reconstruct the video, the compressed data are decoded by inverse processes corresponding to the entropy coding, quantization, transform, and prediction. SUMMARY According to an aspect, a method for encoding or decoding a video is provided. The method comprises determining, for at least one pixel of a block of a picture, a predictor sample based on an intra prediction along a non-diagonal direction, obtaining at least four reference samples of the predictor sample from among a plurality of decoded pixels neighboring the block, obtaining a value for the predictor sample based on an interpolation that uses the at least four reference samples, encoding or decoding the block based on the value. According to another aspect, an apparatus for encoding or decoding a video is provided. The apparatus comprises one or more processors operable to determine, for at least one pixel of a block of a picture, a predictor sample based on an intra prediction along a non-diagonal direction, obtain at least four reference samples of the predictor sample from among a plurality of decoded pixels neighboring the block, obtain a value for the predictor sample based on an interpolation that uses the at least four reference samples, encode or decode the block based on the value. In some embodiments, the interpolation is based on a filter which uses a threshold based linear interpolation. In other embodiments, the interpolation uses an n-tap filter, which is implemented as linear interpolation plus an update term, with n being an integer, for instance 4 or 6. In some embodiment, the interpolation is based on a 4-tap filtering or 6-tap filtering. In a variant, the coefficients of the 4-tap or 6-tap filter are determined once per row of the block. In other variant, coefficients of the 4-tap or 6-tap filter are determined based on a fractional part of an horizontal offset of a position of the predictor sample from the pixel. In other embodiments, the interpolation is based on a linear interpolation and at least one update term. In a variant, the at least one update term is determined using the at least four reference samples and a factor that depends on a fractional part of an horizontal offset of a position of the predictor sample from the pixel. In other variant, the factor is determined once per row of the block. Further embodiments that can be used alone or in combination are described herein. One or more embodiments also provide a computer program comprising instructions which when executed by one or more processors cause the one or more processors to perform the method for encoding or decoding a video according to any of the embodiments described herein. One or more of the present embodiments also provide a non-transitory computer readable medium and/or a computer readable storage medium having stored thereon instructions for encoding or decoding a video according to the methods described herein. One or more embodiments also provide a computer readable storage medium having stored thereon a bitstream generated according to the methods described herein. One or more embodiments also provide a method and apparatus for transmitting or receiving the bitstream generated according to the methods described above. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 illustrates a block diagram of a system within which aspects of the present embodiments may be implemented. FIG.2 illustrates a block diagram of an embodiment of a video encoder within which aspects of the present embodiments may be implemented. FIG.3 illustrates a block diagram of an embodiment of a video decoder within which aspects of the present embodiments may be implemented. FIG.4 illustrates an example of a flowchart for intra prediction of a block in ECM software. FIG.5 illustrates example of reference samples for a block to be predicted. FIG.6 illustrates an example of predictor sample interpolation using six reference samples. FIG.7 illustra