Controlling a video content system
View Patent ↗According to examples of the presently disclosed subject matter, a method of controlling a video content system can include: obtaining a current input frame and a preceding input frame from an input video sequence and obtaining a current degraded frame and a preceding degraded frame from a degraded video sequence corresponding to the input video sequence; computing a first differences value from the current input and the preceding input frames and a second differences value from the current degraded frame and the preceding degraded frame, comparing the first and second differences values, giving rise to an inter-frame quality score; computing an intra-frame quality score using an intra-frame quality measure that is applied in the pixel-domain of the current degraded frame and the current input frame and providing a configuration instruction to the video content system based on a quality criterion related to the inter-frame and the intra-frame quality scores.
1. A method, comprising:
partitioning a current input frame and a current degraded frame according to a tile grid, giving rise to a plurality of tile pairs, wherein the current degraded frame is generated from the current input frame according to a compression parameter value;
calculating a quality score for each tile pair giving rise to a plurality of intra-tile quality scores;
applying perceptual weighting to the plurality of intra-tile quality scores, giving rise to perceptually weighted intra-tile quality scores, including calculating a weight for a per-tile score based on a relation between the per-tile score and at least one other per-tile score; and
pooling the perceptually weighted intra-tile quality scores to compute a quality score for the current degraded frame.
2. The method according to claim 1 , wherein said tile grid is determined based on the frame dimensions.
3. The method according to claim 2 , wherein said tile grid is determined by selecting a tile size that yields at least a pre-determined number of tiles per frame.
4. The method according to claim 2 , wherein said tile grid is determined by selecting a tile size that yields at least a pre-determined number of tiles per tile grid row and/or column.
5. The method according to claim 1 , wherein said tile grid is determined by implementing a tile specification which is compatible with a characteristic of a per tile quality measure that is used for calculating the quality score for each tile pair.
6. The method according to claim 5 wherein said tile specification sets forth a minimum and/or maximum tile size that is compatible with the per tile quality measure.
7. The method according to claim 5 , wherein said tile specification sets forth a tile size that is an integer multiple of a tile dimension that is compatible with the per tile quality measure.
8. The method according to claim 5 , wherein said tile specification sets forth a placement of the tiles in a grid that is aligned with tile locations that are compatible with the per tile quality measure.
9. The method according to claim 1 , wherein said determining a tile grid further includes specifying for at least one tile pair a corresponding sub-tile grid, giving rise to a plurality of sub-tile pairs of the tile pair, and wherein said calculating comprises calculating a quality score for the tile pair wherein at least one component of the quality measure is calculated for each of the plurality of sub-tile pairs, giving rise to multiple sub-tile quality indications, which are then perceptually pooled to yield the intra-tile quality score.
10. The method according to claim 1 , wherein said applying further includes identifying outlier tile scores and assigning an exclusion indication for each of the identified outlier tile scores, and wherein said pooling comprises excluding the indicated outlier tile scores.
11. The method according to claim 1 , wherein said applying further includes calculating a weight for a per-tile score based on a location of a respective tile within the tile grid or in the frame.
12. The method according to claim 1 , wherein said applying further includes calculating a weight for a per-tile score based also on frame characteristics.
13. The method according to claim 9 , wherein said applying perceptual weighting further includes adapting a weight for a given tile according to a relation among sub-tile scores of at least some of the sub-tiles in the tile.
14. The method according to claim 1 wherein said quality score for the frame is further scaled according to characteristics of a respective frame.
15. The method according to claim 1 wherein said quality score for the frame is further scaled according to relations among the per-tile score values.
16. The method according to claim 1 wherein said quality score for the frame is further scaled according to score component values.
17. The method according to claim 9 wherein said quality score for the frame is further scaled according to relations among the per sub-tile quality indications of at least some tiles of the frame.
18. A device, comprising a processor operatively connected to a non-transitory computer readable memory, the processor comprising processing circuitry, the processor configured to:
partition a current input frame and a current degraded frame according to a tile gird, giving rise to a plurality of tile pairs, wherein the current degraded frame is generated from the current input frame according to a compression parameter value;
calculate a quality score for each tile pair giving rise to a plurality of intra-tile quality scores;
apply perceptual weighting to the plurality of intra-tile quality scores, giving rise to perceptually weighted intra-tile quality scores, including calculate a weight for a per-tile score based on a relation between the per-tile score and at least one other per-tile score; and
pool the perceptually weighted intra-tile quality scores to compute a quality score for the current degraded frame.
19. The device according to claim 18 , wherein the processor is configured to partition the current input frame and the current degraded frame further based on the frame dimensions.
20. The device according to claim 19 , wherein the processor is configured to partition the current input frame and the current degraded frame such that there is at least at least a pre-determined number of tiles per tile grid row and/or column.
21. The device according to claim 18 , wherein the processor is configured to implement a tile specification which is compatible with a characteristic of a per tile quality measure that is used by the processor to calculate the quality score for each tile pair.
22. The device according to claim 21 , wherein the tile specification sets forth a minimum and/or maximum tile size that is compatible with the per tile quality measure.
23. The device according to claim 21 , wherein the tile specification sets forth a tile size that is an integer multiple of a tile dimension that is compatible with the per tile quality measure.
24. The device according to claim 21 , wherein said tile specification sets forth a placement of the tiles in a grid that is aligned with tile locations that are compatible with the per tile quality measure.
25. The device according to claim 18 , wherein the processor is further configured to:
specify for at least one tile pair a corresponding sub-tile grid, giving rise to a plurality of sub-tile pairs of the tile pair; and
calculate a quality score for the tile pair, wherein at least one component of the quality measure is calculated for each of the plurality of sub-tile pairs, giving rise to multiple sub-tile quality indications which are then perceptually pooled to yield the intra-tile quality score.
26. The device according to claim 18 , wherein the processor is further configured to identify outlier tile scores and to assign an exclusion indication for each of the identified outlier tile scores, and wherein the processor is adapted to exclude the indicated outlier tile scores.
27. The device according to claim 18 , wherein the processor is further configured to calculate a weight for a per-tile score based on a location of a respective tile within the tile grid or in the frame.
28. The device according to claim 18 , wherein the processor is further configured to calculate a weight for a per-tile score based on frame characteristics.
29. The device according to claim 25 , wherein the processor is further configured to adapt a weight for a given tile according to a relation among sub-tile scores of at least some of the sub-tiles in the tile.
30. The device according to claim 18 , wherein the processor is further configured to apply a factor to one or more of the intra-tile quality scores according to characteristics of a respective frame.
31. The device according to claim 18 , wherein the processor is further configured to apply a factor to one or more of the intra-tile quality scores according to relations among the per-tile score values.
32. The device according to claim 18 , wherein the processor is further configured to apply a factor to one or more of the intra-tile quality scores according to score component values.
33. The device according to claim 25 , wherein the processor is further configured to apply a factor to one or more of the intra-tile quality scores according to relations among the per sub-tile quality indications of at least some tiles of the frame.
34. A non-transitory computer readable storage medium comprising computer readable program code embodied therein for evaluating a quality of a frame, the computer readable program code configured to:
partition a current input frame and a current degraded frame according to a tile grid, giving rise to a plurality of tile pairs, wherein the current degraded frame is generated from the current input frame according to a compression parameter value;
calculate a quality score for each tile pair giving rise to a plurality of intra-tile quality scores; and
apply perceptual weighting to the plurality of intra-tile quality scores, giving rise to perceptually weighted intra-tile quality scores, including calculate a weight for a per-tile score based on a relation between the per-tile score and at least one other per-tile score; and
pool the perceptually weighted intra-tile quality scores to compute a quality score for the current degraded frame.
35. A method, comprising:
partitioning a current input frame and a current degraded frame according to a tile grid, giving rise to a plurality of tile pairs, wherein the current degraded frame is generated from the current input frame according to a compression parameter value, and wherein said tile grid is determined by implementing a tile specification which is compatible with a characteristic of a per tile quality measure that is used for calculating the quality score for each tile pair;
calculating a quality score for each tile pair giving rise to a plurality of intra-tile quality scores;
applying perceptual weighting to the plurality of intra-tile quality scores, giving rise to perceptually weighted intra-tile quality scores; and
pooling the perceptually weighted intra-tile quality scores to compute a quality score for the current degraded frame.