Generalized harmonic curvature as diffusion coefficient for image de-nosing

In this paper, we propose the concept of generalized harmonic curvature for the image surface, and then, we give a new diffusion coefficient which is based on the generalized harmonic curvature for the PM model. Moreover, we also give a semi-implicit difference scheme, the stability with respect to the sup-norm doesn't depend on the time step. Finally, compared with the gauss curvature-driven diffusion model, the proposed model works better even if the noise level is high, a numerical experiment is also given.