No-reference video/image quality measurement with compressed domain features

1. A method for objectively determining perceived video/image quality, the method comprising:

receiving, by a device, a bit-stream comprising encoded video/image data corresponding to an original frame or an original sequence of frames;

parsing the bit-stream to extract a set of video/image coding components, wherein the extracted set of video/image coding components comprise at least one of a set of intra-prediction modes, a set of discrete cosine transform (DCT) coefficients, and motion information; and

performing a prediction function in accordance with the extracted set of video/image coding components prior to decoding the bit-stream to estimate a quality of experience (QoE) score for the encoded video/image data.

2. The method of claim 1 , wherein the extracted set of video/image coding components comprise the motion information, and

wherein performing the prediction function to estimate the QoE score comprises estimating a temporal complexity of the sequence of frames in accordance with the motion information.

3. The method of claim 1 , wherein the extracted set of video/image coding components comprise the set of intra-prediction modes and the set of DCT coefficients, and

wherein performing the prediction function to estimate the QoE score comprises estimating a spatial complexity of an intra-coded frame based at least in part on the set of intra-prediction modes and the set of DCT coefficients.

4. The method of claim 1 , wherein the sequence of frames comprises an intra-coded frame comprising a plurality of macroblocks or blocks,

wherein the extracted set of video/image coding components include a first set of DCT coefficients corresponding to a first one of the macroblocks or blocks, and

wherein performing the prediction function comprises estimating an average luma value for the first macroblock in accordance with the first set of DCT coefficients.

5. The method of claim 1 , wherein the sequence of frames comprises an intra-coded frame comprising a plurality of macroblocks or blocks,

wherein the extracted set of video/image coding components further comprise a plurality of quantization parameter values each of which corresponding to a different one of the macroblocks or blocks, and

wherein performing the prediction function comprises determining an average quantization parameter value across all macroblocks or blocks.

6. The method of claim 1 , wherein the encoded video/image data corresponds to the original

sequence of frames comprising a plurality of individual frames, and

wherein the performing the prediction function comprises:

estimating an individual QoE score for each frame; and

computing a weighted average of all of the individual QoE scores.

7. The method of claim 1 , wherein the QoE score is estimated without reference to the original frame or the original sequence of frames.

8. The method of claim 1 , wherein the extracted set of video/image coding components comprise each of the set of intra-prediction modes, the set of DCT coefficients, and the motion information.

9. A method for objectively determining perceived video/image quality, the method comprising:

receiving a bit-stream comprising encoded video/image data corresponding to an original frame or an original sequence of frames;

parsing the bit-stream to extract a set of video/image coding components, wherein the extracted set of video/image coding components comprise at least one of a set of intra-prediction modes, a set of discrete cosine transform (DCT) coefficients, and a motion information;

determining a content classification corresponding to the original frame or the original sequence of frames in accordance with the extracted set of video/image coding components; and

performing a prediction function to estimate a quality of experience (QoE) score for the encoded video/image data prior to decoding the bit-stream, wherein performing the prediction function comprises estimating the QoE score for the encoded video/image data based at least in part on a known QoE score associated with training data, and wherein the training data has a similar content classification to that of the original frame or the original sequence of frames.

10. The method of claim 9 , wherein the training data comprises a collection of subjective image quality assessments in accordance with human responses to subjective testing.

11. The method of claim 9 , wherein the QoE score is generated without reference to the original frame or the original sequence of frames.

12. A non-transitory computer program product comprising computer executable instructions stored on a non-transitory computer readable medium, the computer executable instructions including instructions to:

receive a degraded bit-stream comprising a plurality of video/image coding components corresponding to a plurality of macroblocks or blocks of an original frame or an original sequence of frames;

parse the degraded bit-stream to extract a set of the video/image coding components;

estimate an average luma value for at least one of the macroblocks or blocks in accordance with the extracted set of video/image coding components; and

estimate a quality of experience (QoE) score corresponding to the degraded video/image bit-stream based at least in part on the average luma value prior to decoding of the degraded bit-stream.

13. The non-transitory computer program product of claim 12 , wherein the extracted set of video/image coding components comprise a set of discrete cosine transform (DCT) coefficients corresponding to the at least one of the macroblocks or blocks.

14. The non-transitory computer program product of claim 12 , wherein the extracted set of video/image coding components comprise a set of intra-prediction modes corresponding to the at least one macroblock.

15. The non-transitory computer program product of claim 12 , wherein the QoE score is generated without reference to the original frame or the original sequence of frames.

16. A non-transitory computer program product comprising computer executable instructions stored on a non-transitory computer readable medium, the computer executable instructions including instructions to:

receive a degraded bit-stream comprising a plurality of video/image coding components corresponding to an original frame or an original sequence of frames;

parse the degraded bit-stream to extract a set of the video/image coding components;

estimate a complexity of the original frame or the original sequence of frames in accordance with the extracted set of video/image coding components; and

estimate a quality of experience (QoE) score corresponding to the degraded video/image bit-stream based at least in part on the complexity of the original frame or the original sequence of frames, the QoE score being estimated prior to decoding of the degraded bit-stream.

17. The non-transitory computer program product of claim 16 , wherein the extracted set of video/image coding components comprise a set of discrete cosine transform (DCT) coefficients corresponding to the original frame or the original sequence of frames.

18. The non-transitory computer program product of claim 17 , wherein the extracted set of video/image coding components comprise motion information corresponding to the original frame or the original sequence of frames.

19. The non-transitory computer program product of claim 17 , wherein the QoE score is generated without reference to the original frame or the original sequence of frames.