Image representation using triangle meshes with explicit discontinuities

A new mesh model for images, which explicitly represents discontinuities (i.e., image edges), is proposed along with a corresponding mesh-generation method that determines the mesh-model parameters for a given input image. The proposed approach is demonstrated to be capable of producing mesh-based image approximations of higher quality (both objectively and subjectively) than those generated by the state-of-the-art error-diffusion (ED) method of Yang, Wernick, and Brankov. For example, the image approximations produced by the proposed method are often about 2.5 dB higher in PSNR than those produced by the ED scheme.