A new algorithm for infrared image restoration based on multi-scale morphological wavelet and Hopfield neural network

Based on the complexity and randomness of the infrared image degradation factors, and integrates the strong de-noising features of multi-scale morphological wavelet and the salient problem solving features of Hopfield neural network in optimization, this paper presents a new algorithm for infrared degraded image restoration. The algorithm takes advantage of the continuous recycle between "multi-scale morphological wavelet de-noising" and "Hopfield neural network iteration" so as to makes access to a better recovery of infrared images. The algorithm also solves the problems in noise suppression and image detail protection of traditional Hopfield neural network image restoration algorithm and successfully protects the edge of the recovery images and details. Simulation results prove the effectiveness of the recovery algorithm.