Hybrid Block-based Motion Estimation Considering Regional Motion Characteristics
Young Duk Kim, Joonyoung Chang, Moon Gi Kang · ITC-CSCC :International Technical Conference on Circuits Systems, Computers and Communications · 2008
In this paper, we propose a hybrid block-based motion estimation algorithm that considers regional motion characteristics. The proposed method combines multiple search algorithms according to regional motion types to produce more accurate motion vector fields while preventing the increase of the computational cost. For these reasons, the proposed method first divides an entire image into three regions with different motion types. Experimental results show that the proposed motion estimation method improves the accuracy of the estimated motion vector fields over the existing algorithms.