A Subpel Motion Vector Estimation for Embedded Video Encoder
Wissal Hassen, Hamid Amiri · International Review on Computers and Software (IRECOS) · 2015
In this work which falls within the overall context of video compression, we are concerned with the elimination of temporal redundancy by motion prediction. The motion vector is the main element of this technique and the precision of estimation of this vector improves the predicted image quality. However, increasing the motion vector precision requires a higher resolution of reference frame. In this paper, we propose a new technique of motion estimation with interpolation by combining the six tap filter and the bilinear interpolation. The proposed technique gives good results compared with predicted images without interpolation.