Video bit stream test
View Patent ↗The present invention is a method for MPEG compliance testing that designs test bit streams such that every frame of the expected decoded video is known. The test bit stream is generated, decoded and stored as an expected output decoded video. The test bit stream is fed to the device under test where it is decoded. Actual output video is compared with the expected decoded video on a frame-by-frame basis. The test results are indicated by discrepancies, which can be visually examined by human or by computer, through a comparison of the two videos. During generation of the test bit stream, for a given GOP structure, the subsequent P and B frames are determined by varying the selected testing MPEG parameter with the whole range of its possible values, while keeping other parameters constant.
1. A method of testing a decoder, comprising the steps of:
generating a data stream;
calculating an expected decoded output from the data stream, the expected decoded output comprising a first plurality of video frames;
decoding the data stream using the decoder to produce a measured output comprising a second plurality of video frames;
generating a plurality of differential video frames, each differential video frame generated as a difference between one of the first plurality of video frames and an associated one of the second plurality of video frames, and
identifying a decoder error from the plurality of differential video frames.
2. The method of claim 1 , wherein the step of generating a data stream comprises the steps of:
selecting an I frame of a Moving Picture Experts Group (MPEG)-formatted video;
determining P frames and B frames subsequent to the I frame by varying an MPEG parameter while keeping other MPEG parameters constant; and
generating the data stream from the I frame, and the P frames and the B frames.
3. The method of claim 2 , wherein the MPEG parameter is a frequency-domain parameter.
4. The method of claim 1 , wherein the step of calculating an expected decoded output from the data stream comprises:
providing the data stream to a debugged decoder.
5. The method of claim 1 , wherein the step of calculating an expected decoded output from the data stream comprises:
computing the data stream using a software decoder.
6. The method of claim 1 , wherein the step of identifying the decoder error from the plurality of differential video frames comprises the step of:
identifying a non-black pixel as a decoder error.
7. The method of claim 6 , further comprising the step of:
determining the decoder error from where and when the non-black pixels occur.
8. An apparatus for testing a decoder, comprising:
means for generating a data stream;
means for calculating an expected decoded output from the data stream, the expected decoded output comprising a first plurality of video frames;
means for providing the data stream to the decoder to produce an measured output comprising a second plurality of video frames;
means for generating a plurality of differential video frames, each differential video frame generated as a difference between one of the first plurality of video frames and an associated one of the second plurality of video frames; and
means for identifying a decoder error from the plurality of differential video frames.
9. The apparatus of claim 8 , wherein the means for generating a data stream comprises the steps of:
means for selecting an I frame of a Moving Picture Experts Group (MPEG)-formatted video;
means for determining P frames and B frames subsequent to the I frame by varying an MPEG parameter while keeping other MPEG parameters constant; and
generating the data steam from the I frame, and the P frames and the B frames.
10. The apparatus of claim 9 , wherein the MPEG parameter is a frequency-domain parameter.
11. The apparatus of claim 8 , wherein the means for calculating an expected decoded output from the data steam comprises a debugged decoder.
12. The apparatus of claim 8 , wherein the means for calculating an expected decoded output from the data stream comprises a software decoder.
13. The apparatus of claim 8 , wherein the means for identifying the decoder error from the plurality of differential video frames comprises the step of:
means for identifying a non-black pixel as a decoder error.
14. The apparatus of claim 13 , further comprising means for determining the decoder error from where and when the non-black pixels occur.
15. The method of claim 2 , wherein the MPEG parameter is varied to the whole range of its possible values.
16. The apparatus of claim 8 , wherein the MPEG parameter is varied to the whole range of its possible values.