Adaptive search range determination for geometry based disparity and motion estimation of MVC

Multi-view video consists of a set of multiple video sequences from multiple viewpoints or view directions in the same scene. It contains extremely a large amount of data and some extra information to be stored or transmitted to the user and computational complexity of processing. In order to achieve a better coding performance in multi-view video coding (MVC), the search range for both motion estimation (ME) and disparity estimation (DE) are set twice as large as necessary for ME alone, which brings great computational complexity to the encoder. Search range (SR) of motion search constraints current block to search the matched block within a predefined area in the reference picture. This research proposed an algorithm to solve the redundancy reduction problem to get fast disparity and motion estimation for reducing complexity. We use geometry and adaptive search range by adjust macroblock (MB) type that suitable with window SR. Experimental results show that the proposed method can reduce the encoding time, so that the encoder complexity can be reduced but achieves almost the same compress quality. The encoding time reduce up more than 70%.