The Research of Motion Estimation in the Field of Video Compression
|School||Beijing University of Chemical Technology|
|Course||Applied Computer Technology|
|Keywords||Video Coding Motion Estimation Block-matching Correlation Center bias|
Video coding is one of the hot topics in the field of information technology today , due to a strong correlation of the video sequence of images in the timeline , motion estimation (ME) and motion compensation (MC) technology can effectively reduce the time , so the technology has been widely used in a variety of video compression coding schemes such as MPEG and H.26x . This paper first introduces some basic technology of motion estimation , block size , starting point forecast , search accuracy and search strategies , and simulation of several block -matching motion estimation algorithms and their performance analyzed . For block matching motion estimation algorithm exist some defects , such as full use of the adjacent vector correlation and center-biased characteristic proposed two improved motion estimation algorithm for video coding movement , respectively, as a estimated new algorithm (NITSS) the diamond search algorithm based on adaptive prediction (APDS) . These two algorithms full use of the correlation of the motion vector and the center bias characteristic was improved on this basis for the choice of the search starting point , the search model and search strategies . The simulation results show that , compared with the existing algorithm , such as three-step search method, four-step search , diamond search , these two algorithms in search accuracy and search time there is a certain improvement .