A New Method for Noise Elimination Using Fuzzy Neural Network Optimized by Genetic Algorithm
Sajjad Ghatei · 2012
In common applications source of image information are required for analyzing and description. When applied image changes from special form in to another such as: storage, digitalizing, some distortion may occur in image. So, output image needs optimization. For this purpose, there are some algorithms and different techniques with special advantages, disadvantages and restrictions. In this paper, different types of classic spatial filters like: median filter, Multilevel FIR-Median Hybrid Filter and fuzzy neuron networks filters have been discussed and suggested a new fuzzy neuron network optimized by genetic algorithm to improve missing point of image. For better result, we trained fuzzy neuron network with an artificial image. The results showed fuzzy neuron network filter optimized by genetic algorithm (PSNR=32.5697, 28.9043, 31.0041) had better results in comparison with another techniques.