A Novel Frame Rate Conversion Algorithm Based on Adaptive Motion Search

摘要

In this work, we propose a new frame rate conversion algorithm (FRCA) based on adaptive motion estimation and compensation. In the proposed algorithm, an adaptive diamond strategy is firstly used to find out all the motion vectors for each block by adaptively applying diamond algorithm in the current frame. Secondly, the motion vector for each block is further refined by full search algorithm to reduce the edge blur and the motion blur caused by fast moving objects in the interpolated frames. The proposed algorithm achieves a big improvement in terms of image quality and greatly reduces computational complexity compared to conventional motion estimation algorithms such as full search algorithm and diamond search algorithm.