Method and apparatus for detecting quality defects in a video bitstream

Patent metadata

Field	Value
Publication number	US-9686515-B2
Application number	US-201114363607-A
Country	US
Kind code	B2
Filing date	Dec 9, 2011
Priority date	Dec 9, 2011
Publication date	Jun 20, 2017
Grant date	Jun 20, 2017

Abstract

There is provided an apparatus for detecting a quality defect in a video sequence, the apparatus comprising: a receiver arranged to receive a video bitstream representing a video sequence; and a defect module arranged to determine a measure of quantization parameter for each picture of the video sequence, the defect module further arranged to identify a quality defect when an abrupt change in the measure of quantization parameter occurs.

First claim

The invention claimed is:

1. A method for detecting a quality defect in a video sequence, the method comprising:
receiving, by a receiver, an encoded video bitstream representing a video sequence;
determining, by the receiver, a measure of quantization parameter (QP) for each picture of the video sequence; and
identifying, by the receiver, without decoding the encoded video bitstream, a quality defect when an abrupt change in the measure of QP occurs, wherein the abrupt change in the measure of QP is detected when a first picture in the video sequence has a measure of QP that i) is greater than a measure of QP for a preceding picture by more than a first threshold t 1 and ii) is greater than a measure of QP for a subsequent picture by more than a second threshold t 2 .

2. The method of claim 1 , further comprising:
determining if each picture of the video sequence is an intra-coded picture; and
identifying a quality defect as intra-picture flicker when the abrupt change in the measure of the quantization parameter (QP) occurs such that the first picture is an intra-coded picture and the preceding picture and the subsequent picture are both inter-coded pictures.

3. The method of claim 2 , further comprising determining whether the first picture is part of a scene change, wherein the quality defect is detected further in response to the determination that the first picture is not part of a scene change.

4. The method of claim 1 , wherein the first threshold t 1 and the second threshold t 2 which are used to detect the abrupt change in the measure of quantization parameter are both equal to a threshold amount, and the abrupt change is detected when the measure of quantization parameter for the first picture is greater than a measure of quantization parameter of a first predetermined number of preceding pictures and is greater than a second predetermined number of subsequent pictures.

5. The method of claim 4 , wherein the first predetermined number of pictures is equal to the second predetermined number of pictures.

6. The method of claim 5 , wherein the predetermined number of pictures is one.

7. The method of claim 1 , further comprising:
determining an average motion vector; and
generating a quality score based on identifying the quality defect and the average motion vector.

8. The method of claim 1 , wherein the measure of quantization parameter for a picture is the average quantization parameter used for the picture.

9. The method of claim 1 , wherein the method is used in a quality estimation model to determine a perceptual quality of the video sequence represented by the video bitstream.

10. The method of claim 9 , wherein the perceptual quality of the video sequence is determined prior to any decoding of the encoded video bitstream.

11. The method of claim 9 , wherein the method is used in a node in a distribution network different from both a source of the encoded video bitstream and an end-user destination of the encoded video bitstream.

12. The method of claim 1 , wherein the video bitstream is encoded using at least one of: H.264, MPEG-2, H.263, MPEG-4 and HEVC/H.265.

13. The method of claim 1 , wherein the preceding picture immediately precedes the first picture and the subsequent picture immediately follows the first picture.

14. The method of claim 1 , further comprising generating a quality score based on identifying the quality defect and based on at least one of: a) a number of intra-picture flicker occasions; b) frequency of intra-picture flicker; c) distance, in terms of time or number of pictures, between intra-picture flicker events; and d) duration of drop in encoding quality starting at an intra-picture.

15. The method of claim 1 , further comprising estimating quality of the video sequence based on identifying the quality defect and based on a resolution of the video sequence.

16. The method of claim 1 , wherein the encoded video bitstream is block-coded such that each picture of the video sequence includes a plurality of blocks, each block having a QP value and wherein determining the measure of QP for each picture of the video sequence comprises averaging the QP value of each block of the plurality of blocks of the picture.

17. The method of claim 1 , further comprising:
determining a QP range for the video sequence, wherein determining the QP range for the video sequence comprises calculating QPmax−QPmin, wherein QPmax is the maximum QP value for the video sequence and QPmin is the minimum QP value for the video sequence; and
calculating a quality score for the video sequence using a quality model, the determined QP range, and a defect indicator indicating that the quality defect has been identified.

18. The method of claim 1 , further comprising:
obtaining a motion indication indicative of the motion of the video sequence; and
calculating a quality score for the video sequence using a quality model, the obtained motion indication, and a defect indicator indicating that the quality defect has been identified.

19. The method of claim 18 , wherein
the method further comprises determining a QP range for the video sequence, wherein determining the QP range for the video sequence comprises calculating QPmax−QPmin, wherein QPmax is the maximum QP value for the video sequence and QPmin is the minimum QP value for the video sequence, and
the step of calculating the quality score for the video sequence comprises calculating the quality score using the quality model, the obtained motion indication, the determined QP range, and the defect indicator indicating that the quality defect has been identified.

20. An apparatus for detecting a quality defect in a video sequence, the apparatus comprising:
a receiver arranged to receive an encoded video bitstream representing a video sequence; and
a defect module arranged to determine a measure of quantization parameter for each picture of the video sequence, the defect module further arranged to identify, without decoding the encoded video bitstream, a quality defect when an abrupt change in the measure of QP occurs, wherein the abrupt change in the measure of QP is detected when a first picture in the video sequence has a measure of QP that i) is greater than a measure of QP for a preceding picture by more than a first threshold t 1 and ii) is greater than a measure of QP for a subsequent picture by more than a second threshold t 2 .

21. The apparatus of claim 20 , further comprising:
a picture type determining module arranged to determine whether each picture of the video sequence is an intra-coded picture; and
wherein the defect module is further arranged to identify a quality defect as an intra-picture flicker when the abrupt change in the measure of quantization parameter occurs such that the first picture is an intra-coded picture and the preceding picture and the subsequent picture are both inter-coded pictures.

22. The apparatus of claim 20 , further comprising a quality module arranged to receive the video bitstream and a defect indication from the defect module, the quality module further arranged to apply a quality model to estimate a quality of video reconstructed from the video bitstream.

23. The apparatus of claim 22 , wherein the quality module is further arranged to generate a quality score based on the defect module identifying the quality defect and based on at least one of:
the level of spatial detail in a picture or scene in the video sequence; and
the amount

Assignees

• Pettersson Martin
• Argyropoulos Savvas
• ERICSSON TELEFON AB L M (publ)

Inventors

• Pettersson Martin
• Argyropoulos Savvas

Classifications

• H04N17/004 Primary

  for digital television systems · CPC title

• H04N17/04 Primary

  for receivers · CPC title

• H04N19/89

  involving methods or arrangements for detection of transmission errors at the decoder · CPC title

• H04N21/44008

  involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream (arrangements characterised by components specially adapted for monitoring, identification or recognition of video in broadcast systems H04H60/59) · CPC title

• H04N19/85 Primary

  using pre-processing or post-processing specially adapted for video compression · CPC title