Scene change detection for perceptual quality evaluation in video sequences

Patent metadata

Field	Value
Publication number	US-10349048-B2
Application number	US-201114351483-A
Country	US
Kind code	B2
Filing date	Oct 11, 2011
Priority date	Oct 11, 2011
Publication date	Jul 9, 2019
Grant date	Jul 9, 2019

Abstract

There are provided a plurality of methods for detecting a scene change in a streamed video, the streamed video comprising a series of pictures. An example method comprises calculating, for a plurality of positions, a difference between the costs of coding macro-blocks at the same position in successive pictures. The method further comprises identifying a new scene when the sum of the differences for a plurality of positions meets a threshold criterion. There is further provided a method of determining the perceptual impact of a packet loss on a streamed video the method comprising: identifying a packet loss; and determining if the lost packet contained information relating to a picture at the start of a new scene, wherein a new scene is detected using one of the methods disclosed herein.

First claim

The invention claimed is:

1. A method of detecting a scene change in a streamed video, the streamed video comprising a series of pictures, the method comprising:
calculating, for a plurality of positions, a difference between the costs of coding macro-blocks at the same position in successive pictures, wherein the successive pictures includes a first picture and a second picture immediately following the first picture in the series of pictures;
summing the calculated differences between the costs of coding macro-blocks at the same position in successive pictures to produce a total difference;
determining that the total difference for a plurality of positions meets a threshold criterion; and
identifying a new scene as a result of determining that the total difference for a plurality of positions meets the threshold criterion,
wherein a new scene is identified when the total difference for a plurality of pixel positions is either higher than a maximum threshold value or lower than a minimum threshold value, and
wherein at least one threshold value is determined by taking a weighted average of differences calculated between a predetermined number of pictures that are immediately preceding the first picture and the second picture, wherein the differences between the pictures that are more recently preceding the first picture and the second picture are given a greater weight.

2. The method of claim 1 , wherein the cost of coding a macro-block is determined according to the number of bits used to encode the macro-block.

3. The method of claim 1 , wherein the cost of coding a macro-block is determined according to the motion vectors used to encode the macro-block.

4. The method of claim 1 , wherein the cost of coding a macro-block is determined according to the quantization parameter value used to encode the macro-block.

5. The method of claim 1 , wherein the cost of coding a macro-block is determined according to macro-block type.

6. The method of claim 5 , wherein the macro-block type comprises one of: skip; inter-coded macro-block as one partition; inter-coded macro-block subdivided into multiple partitions; intra-coded macro-block as one partition; and intra-coded macro-block subdivided into multiple partitions.

7. The method of claim 1 , further comprising applying motion compensation, and wherein the difference between the cost of coding macro-blocks for pixels at the same position in successive pictures is calculated for the pixel positions after motion compensation has been applied.

8. The method of claim 1 , further comprising:
including an offset between the positions in successive pictures;
for each of a plurality of offset values, calculating a sum of the differences for a plurality of positions to obtain a plurality of difference totals corresponding to the plurality of offset values; and
comparing the minimum of the plurality of difference totals to the threshold criterion.

9. The method of claim 1 , wherein at least one threshold value is determined from the differences calculated between successive pictures of a plurality of preceding pictures.

10. The method of claim 1 , further comprising selecting the plurality of positions to be compared between successive pictures.

11. The method of claim 10 , wherein selecting the plurality of positions comprises selecting every position.

12. The method of claim 10 , wherein selecting the plurality of positions comprises selecting a predetermined subset of positions.

13. The method of claim 10 , wherein selecting the plurality of positions comprises selecting a random pattern of positions.

14. The method of claim 1 , wherein the difference between the costs of coding macro-blocks at the same position in successive pictures is calculated by taking the minimum value of differences between the costs of coding macro-blocks at a plurality of positions offset from the same position in successive pictures.

15. A non-transitory computer-readable medium, carrying instructions, which, when executed by computer logic, causes said computer logic to carry out any of the methods defined by claim 1 .

16. An apparatus for detecting a scene change in a streamed video, the streamed video comprising a series of pictures, the apparatus comprising:
a hardware implemented processor arranged to calculate, for a plurality of positions, a difference between the cost of coding macro-blocks at the same position in successive pictures, wherein the successive pictures includes a first picture and a second picture immediately following the first picture in the series of pictures;
the processing being further arranged to sum the calculated differences between the costs of coding macro-blocks at the same position in successive pictures to produce a total difference;
the processing being further arranged to determine that the total difference for a plurality of positions meets a threshold criterion; and
a scene change detector arranged to identify a new scene as a result of determining that the total difference for a plurality of positions meets the threshold criterion,
wherein the threshold criterion is determined by taking a weighted average of differences calculated between a predetermined number of pictures that are immediately preceding the first picture and the second picture, wherein the differences between the pictures that are more recently preceding the first picture and the second picture are given a greater weight.

17. The apparatus of claim 16 , the processing being further arranged to select the plurality of positions to be compared between successive pictures.

18. The apparatus of claim 17 , wherein selecting the plurality of positions comprises selecting every position.

19. The apparatus of claim 17 , wherein selecting the plurality of positions comprises selecting a predetermined subset of positions.

20. The apparatus of claim 17 , wherein selecting the plurality of positions comprises selecting a random pattern of positions.

21. The apparatus of claim 16 , wherein the difference between the costs of coding macro-blocks at the same position in successive pictures is calculated by taking the minimum value of differences between the costs of coding macro-blocks at a plurality of positions offset from the same position in successive pictures.

Assignees

• Pettersson Martin
• Argyropoulos Savvas
• Lindegren David
• List Peter
• Ericsson Telefon Ab L M
• Deutsche Telekom Ag

Inventors

• Pettersson Martin
• Argyropoulos Savvas
• Lindegren David
• List Peter

Classifications

• H04N5/147 Primary

  Scene change detection · CPC title

• H04N17/004 Primary

  for digital television systems · CPC title

• H04N5/93

  Regeneration of the television signal or of selected parts thereof · CPC title