Applications of transiently chaotic neural networks to image restoration

Abstract

Transiently chaotic neural network with continuous neural states is implemented to restore gray level images. The neural network is modeled to represent the image whose gray level function is the simple sum of the neuron state variables. The restoration consists of two phases: parameter estimation and image reconstruction. During the first phase, parameters are estimated by comparing the energy function of the neural network to a constraint error function. The neural network is updated using stochastic chaotic simulated annealing. Hopfield neural network is also implemented to compare the results. Experiments show that transiently chaotic neural network could get good results in much shorter time compared to Hopfield neural network.