A Fast Adaptive Algorithm for Sub-Pixel Motion Estimation

Real-time video encoding has posed several challenges to the research community. One of them is the design and development of motion estimation algorithms in the sub-pixel domain. Block Matching Algorithms (BMAs) are popular within the video coding standards. Hence, the current work develops iterative and adaptive algorithm for block-based motion estimation. The proposed work makes use of a new adaptive algorithm in integer domain and further goes for fractional pixel search strategy. In the sub-pixel range, it exploits the direction of motion which is considered to be transferred from integer domain to the next half-pixel domain. The computation in the quarter-pixel domain contributes to the robustness of the algorithm.