Self-organizing maps based 3D normal mesh compression

Due to the widespread use of 3D surface meshes in various areas such as computer graphics, animations, virtual reality and visualization, the essential need for an efficient 3D mesh compression technique become more and more crucial. This paper introduces a self organizing feature maps based progressive compression technique for 3D semi-regular meshes. The proposed technique is based on multiresolution decomposition which obtained by the wavelet transform and then it quantize the coefficients by the self organizing feature maps (SOFM) as a vector quantizer which improves the visual quality of the reconstructed mesh. Our experiments show that the proposed technique over performs the state-of-art techniques in Terms of the visual quality of compressed meshes.