Authors: Semih Çelik, Uluğ Bayazıt
Abstract: Visibility computations are commonly used in computer graphics applications. This paper presents a new view-dependent compression technique for 3D animated meshes. The approach consists of geometry coding of visible parts and connectivity coding of changes in visible regions. The proposed view-dependent predictive compression method yields significant improvement in compression performance over compression method without visibility awareness with gains up to 47 %.