Fast Block Matching Criterion for Real-Time Video Communication
Nijad A-Najdawi · 2017
Recently several hardware and software solutions have been provided to transmit videos over low bit-rate channels. Although video decoders are simple and can be easily implemented; however, high quality video encoders for real-time applications are still considered an open area of research for most of the compression algorithms available in the field. In video encoders, full-search motion estimation algorithms results in global optima at the cost of intensive computations that restricts its use in real-time applications. With a trade-off between quality and complexity, various algorithms have been proposed in literature for fast search motion estimation, with the majority of them developed in the spatial domain. For this purpose, researchers have proposed many approaches that perform matching of pixels in consequent video frames using different searching methods. Although those algorithms reduce the total number of required mathematical operations per block when compared to Full Search, the results converge to local minima, and a significant amount of computations is still required. Therefore, in this research, a real-time video compression system has been developed to provide a solution to this problem, and make real-time multimedia applications such as video conferencing affordable. In particular, this paper proposes a motion estimation algorithm in the frequency domain, where the new block matching method examines the similarities between a subset of frequencies in corresponding blocks. This approach significantly reduces both the number of comparisons and the total mathematical computations required per block.