Video coding based on true motion estimation

A new approach for MC-DCT (motion compensated-discrete cosine transform) hybrid video coding based on true motion estimation is proposed. The true motion estimation technique employs the least-median-squares (LMedS) matching criterion in block matching process for motion estimation. The rationale for using such true motion estimator is that at low bit-rates, very few bits are available for coding the motion-compensated DCT errors, and thus the detection of the true underlying motion of the picture block becomes much more important than just trying to find a block-match that satisfies the traditional least-mean-squares (LMS) matching criterion. Without detecting the true motion of the picture block, the errors due to insufficient bits for coding the motion-compensated DCT errors will be propagated to succeeding inter-frames, resulting in a progressive degeneration on perceptual quality of the pictures. An important advantage of using true motion estimation is that it results in better perceptual video quality compared to the conventional block-matching algorithm that utilizes least-mean-squares matching criterion. Experimental results obtained had supported our proposition on using the LMedS-based matching criterion in block-matching motion estimation for low bit rate video coding.

[1]  Olivier Verscheure,et al.  Adaptive quantization using a perceptual visibility predictor , 1997, Proceedings of International Conference on Image Processing.

[2]  Andrea Basso,et al.  Perceptual bit allocation for MPEG-2 CBR video coding , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[3]  Sugato Chakravarty,et al.  Methodology for the subjective assessment of the quality of television pictures , 1995 .

[4]  Itu-T Video coding for low bitrate communication , 1996 .

[5]  Peter J. Rousseeuw,et al.  Robust Regression and Outlier Detection , 2005, Wiley Series in Probability and Statistics.

[6]  P. Holland,et al.  Robust regression using iteratively reweighted least-squares , 1977 .

[7]  Anil K. Jain,et al.  Displacement Measurement and Its Application in Interframe Image Coding , 1981, IEEE Trans. Commun..

[8]  James Hu,et al.  DVQ: A digital video quality metric based on human vision , 2001 .

[9]  Mohammed Ghanbari,et al.  A multi-metric objective picture-quality measurement model for MPEG video , 2000, IEEE Trans. Circuits Syst. Video Technol..

[10]  Zhenghua Yu,et al.  Vision-model-based impairment metric to evaluate blocking artifacts in digital video , 2002, Proc. IEEE.

[11]  Michael Spann,et al.  Robust Optical Flow Computation Based on Least-Median-of-Squares Regression , 1999, International Journal of Computer Vision.

[12]  P. Rousseeuw Least Median of Squares Regression , 1984 .

[13]  Stefan Winkler,et al.  Perceptual distortion metric for digital color video , 1999, Electronic Imaging.

[14]  A. Murat Tekalp,et al.  Digital Video Processing , 1995 .