A total variation model based on edge adaptive guiding function for remote sensing image de-noising

The unexpected noise generated during the process of remote sensing images formation and transmission process is a main factor undermining the images’ quality and usage. In recent years, thanks to its local self-adapting characteristics, formal normalization, and modeling flexibility, PDE has received wide attention for its image de-noising functions, thus pushing the realization of maintaining image details while successfully de-noising a new goal for remote sensing images filtering. Having firstly analyzed and discussed the TV model and M model, a modified variation-model (S model for short) based on edge adaptive guiding function is proposed in this paper. The model introduces edge adaptive guiding function based on the standard gradient into the non-linear diffusion term and re-constructed approaching term, which adaptively adjust the smooth intensity around edge and texture information-rich regions of remote sensing images. S-model does not only overcome staircase effect that is easily produced in the TV model, but also avoids losing details and texture information which is often seen in M model, it can efficiently eliminate noises, maintain a good image edge and keep texture details perfectly. The experimental results validate the effectiveness and stability of the proposed model.