Variable size block motion estimation

The success of video compression standards H.261/H.263, MPEG-1/MPEG-2 is greatly due to fairly efficient motion detection, estimation and compensation stages. In this study, we attempt to improve the inherent search process of the motion estimation and the compensation stage in some application specific cases including highway traffic video sequences. The proposed variable block size motion estimation approach takes advantage of the highly correlated spatial motion in highway traffic sequences for fast yet efficient motion estimation. The methodology presented outperforms the modified logarithmic search (MLS) and the conjugate directions search (CDS) algorithm in reduced complexity and residual entropy.

[1]  Hsueh-Ming Hang,et al.  An efficient block-matching algorithm for motion-compensated coding , 1987, ICASSP '87. IEEE International Conference on Acoustics, Speech, and Signal Processing.

[2]  Graham R. Martin,et al.  Variable size block matching motion estimation with minimal error , 1996, Electronic Imaging.

[3]  M. GHANBARI,et al.  The cross-search algorithm for motion estimation [image coding] , 1990, IEEE Trans. Commun..

[4]  K. R. Rao,et al.  Motion Compensated Interframe Image Prediction , 1985, IEEE Trans. Commun..

[5]  Mohammed Ghanbari,et al.  The Cross-Search Algorithm for Motion Estimation , 1990 .

[6]  Michael W. Marcellin,et al.  Variable size finite-state motion vector quantization , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[7]  Anthony G. Constantinides,et al.  Low bit rate video coding using variable block size model , 1990, International Conference on Acoustics, Speech, and Signal Processing.

[8]  John W. Woods,et al.  Motion vector quantization for video coding , 1995, IEEE Trans. Image Process..