Image Denoising Algorithm Based On Wavelet And Pulse Coupled Neural Network

With the development of computer technology,image processing technology has been widespread concerned and researched.Image in transmission and acquisition process will be affected by noise,it seriously affects the quality of the image and brings great difficulty to image processing.Therefore,image denoising is still a hot topic of current research.On the basis of analyzing the performance of existing image denoising algorithms,the image denoising algorithm based on Wavelet and Pulse Coupled Neural Network(PCNN)is researched,and the main research contents and innovative work are as follows:1.The noise type,the traditional image denoising algorithm,and the principle and working process of the pulse coupled neural network(PCNN)are analyzed,the traditional image denoising algorithms are simulated and analyzed.2.On the basis of analyzing the speckle noise characteristics and studying adaptive threshold denoising algorithm,morphological wavelet threshold denoising algorithm,and Wiener filtering and wavelet fusion denoising algorithm,after PCNN,Wiener filtering,and wavelet denoising algorithm are combined,Wiener filtering and PCNN wavelet based denoising algorithm is peoposed and verified by simulation experiment.3.The anisotropic diffusion image denoising algorithm uses gradient information for edge detection,if the edge of the image is contaminated by noise,this method can not effectively detect the edge of the image,the edge of the image will be destroyed.To solve this problem,on the basis of combining pulse coupled neural network(PCNN)with image entropy,PCNN and image entropy based improved anisotropic diffusion denoising algorithm(PCNN-IEAD)is proposed.This proposed algorithm uses PCNN to calculate the entropy of image sequence,which is used as another edge detection operator and combined with anisotropic diffusion model,avoid being affected by noise duo to gradient edge detection method,and effectively control noise and protect the edge and texture.