Core image inpainting based on neighborhood pixels and wavelet de-noising

Abstract:

In order to repair the broken core image，this paper proposes an inpainting algorithm based on neighborhood pixels and wavelet de-noising. Firstly，the priorities of the pixels which need repairing are calculated in order to determine the filling and inpainting order；then，the pixels are repaired with the neighboring information and the improved weighting function；finally，the de-noising of the restored core image is performed by wavelet de-noising. The algorithm combines image inpainting and wavelet de-noising，and obtains excellent results in repairing experiments.