BP Neural Network for Image Restoration Based on Grey Wolf Optimization Algorithm

A BP neural network image restoration algorithm optimized by gray wolf algorithm is proposed. The three-layer BP neural network can approximate any complex nonlinear relationship, eliminate the high-dimensional equation calculation and the prior generalized stationary hypothesis constraint of the traditional algorithm. The global search ability of the gray wolf optimization algorithm makes up for the shortcomings of the BP neural network performance over-reliance on the initial parameter selection. And the experimental results show that the algorithm has faster convergence speed and higher recovery precision than Wiener filter algorithm and GA-BP algorithm.