A NOVEL MOTION ESTIMATION METHOD FOR MESH-BASED VIDEO MOTION TRACKING

This paper presents a novel motion estimation method for mesh-based video motion tracking. The proposed method has been called mesh-based square-matching (MB-SM) motion estimation method. The MB-SM method outperforms the commonly used motion estimation methods in terms of computational cost reduction, efficiency and image quality (i.e. peak signal to noise ratio (PSNR)). It yields comparable PSNR values with the hexagonal matching method [1] while it has lower computation complexity. 1