Hexagon-Based Motion Estimation using Genetic Algorithm

• Main Authors: Wang-shu Cheng, 鄭婉淑
• Other Authors: Jyh-Horng Jeng
• Format: Others
• Language: en_US
• Published: 2007
• Online Access: http://ndltd.ncl.edu.tw/handle/24745289155779655821

碩士 === 義守大學 === 資訊工程學系碩士班 === 95 === With the improvement of science and technology, the development of the network, and the exploitation of the HDTV, the demands of audio and video become more and more important. Depending on the video coding technology would be the solution for achieving these requirements. Motion estimation, which removes the redundancy in video frames, plays an important role in the video coding. Therefore many experts devote themselves into the issues. The existing fast algorithms relay on the assumption that the matching error decreases monotonically as the searched point moves closer to the global optimum. However, Genetic Algorithm is not fundamentally limited to this restriction. The character would help the proposed scheme to search the mean square error more close to the algorithm of Full Search than those fast algorithms. The proposed technique focuses on combing the Hexagon-based search algorithm, which is faster than Diamond Search, and Genetic Algorithm. Experiments are performed to demonstrate the encoding speed and accuracy of Hexagon-based search pattern method and proposed method.