A Motion Vector Prediction Scheme for MPEG-2 to H.264 Transcoding Based on Smoothness of Motion Vector Field

In MPEG-2 to H.264 transcoding, MPEG-2 motion vector (MV) is often reused as search center to simplify H.264 motion estimation (ME) module. In this paper, we show that motion vector predictor (MVP) from neighboring sub-blocks is more accurate than MPEG-2 MV as search center when MPEG-2 MV field is non-smooth. A criterion based on relative motion is proposed to evaluate smoothness of MPEG-2 MV field. And a hardware oriented MV prediction scheme is also proposed. Experiment results show that the proposed MV prediction scheme with a relative small search range can approach the performance of full search algorithm. Comparing with the method only utilizing MPEG-2 M V, the proposed approach can achieve significant improvement on accuracy of motion prediction, especially in sequences with fast motion and complicate background.

[1]  Ming-Ting Sun,et al.  Digital Video Transcoding , 2005, Proceedings of the IEEE.

[2]  Yu Sun,et al.  Video transcoding: an overview of various techniques and research issues , 2005, IEEE Transactions on Multimedia.

[3]  Jill M. Boyce,et al.  Fast mode decision and motion estimation for H.264 with a focus on MPEG-2/H.264 transcoding , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[4]  Liang-Gee Chen,et al.  Analysis and architecture design of variable block-size motion estimation for H.264/AVC , 2006, IEEE Transactions on Circuits and Systems I: Regular Papers.

[5]  Zhi Zhou,et al.  Fast variable block-size motion estimation algorithm based on merge and slit procedures for H.264/MPEG-4 AVC , 2004, 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512).

[6]  Oscar C. Au,et al.  Enhanced predictive motion vector field adaptive search technique (E-PMVFAST)-based on future MV prediction , 2005, 2005 IEEE International Conference on Multimedia and Expo.

[7]  Zhi Zhou,et al.  Motion information and coding mode reuse for MPEG-2 to H.264 transcoding , 2005, 2005 IEEE International Symposium on Circuits and Systems.