Sorting-based Motion Estimation Algorithm by Using Rough Average Value Calculation
Yue Liu, Hyo-Moon Cho, Sang‐Bock Cho · ITC-CSCC :International Technical Conference on Circuits Systems, Computers and Communications · 2009
A sorting-based algorithm by using RAVC is proposed to reduce the computation cost of block matching for motion estimation. This new algorithm is based on the characteristic of image that pixel matching errors with similar magnitude tend to appear in clusters for natural video sequences. By using this clustering characteristic, we proposed an adaptive Partial Distortion Search algorithm which significantly improved the computation efficiency. And to further reduce the computational complexity, we made use of virtual average value and defined Rough Average Value Calculation (RAVC) method. Simulation results showed that the new algorithm had a significant computational speed-up without introducing any loss in predicted image when compared to conventional PDE and NPDE.