Rate–quality tradeoff MPEG video encoder

摘要

This paper will address a video coding scheme which controls the video subjective quality with bit-rate constraint. In the paper, a video quality measure based on block-based features which has a high correlation with subjective test experiments is presented first. Four block-based features (the average of DFT differences, the standard deviation of DFT differences, the mean absolute deviation of wepstrum differences, and the variance of UVW differences) are extracted from the input and output video sequences and fed into a four-layer backpropagation neural network that has been trained by subjective testing. These features were selected out of a much larger set of candidate features, based on their combined ability to predict observer assessments of image quality. The quality measure is then introduced into the design of a rate–quality tradeoff MPEG encoder which is based on the non-convex quality and bitcount functions at the macroblock level. The relationships between one macroblock's quality, a frame's quality and overall quality have also been addressed. The proposed rate–quality tradeoff scheme has a higher, and also a more consistent, quality measure than the MPEG2 Test Model 5 scheme. The objective quality, as measured by the signal-to-noise ratio is also improved, and the variations in the signal-to-noise ratio and bit-rate from frame to frame are reduced. The results have also been validated by a subjective test experiment. The proposed rate–quality tradeoff scheme opens a different video encoding arena compared to the traditional rate–distortion encoding scheme.