A slope-based search technique for block motion estimation

Motion estimation is the most computationally demanding part of MPEG coding system and it could significantly affect the output quality of an encoded sequence. Thus a new fast motion estimation algorithm is presented, named as the slope-based search technique (SBST), which is actually based on selective full search and aggressive diamond search methods. The SBST algorithm shows a significant improvement over the previous other algorithms in both terms of visual quality and computational complexity. By using a new biased criterion and some additional steps of slope analysis, our simulation results show that the proposed algorithm can reduce the computational complexity with the diamond search algorithm, while proving superior and more robust quality similar to that of the full search algorithm.