IMAGE NOISE REDUCTION Wavelet Transform And ROF Model

Noise removal is a long-standing problem in the field of image processing and image analysis, in the image restoration of the classic problems of the field of study. In addition to traditional filtering denoising methods, wavelet transform and partial differential equations are mathematical methods of the past two decades more popular image processing. Wavelet transform has a frequency localized multi-scale analysis of the nature of the de-noising effect is obvious, but easy fuzzy edge the partial differential equations can protect and reduce noise while enhancing image edges and details, but the time complexity is high. How to select these two methods in practical applications, and how to play the advantages of each of these two methods is the purpose of this study.This paper analyzes the wavelet transform and partial differential equations ROF model diffusion behavior and edge protection mechanism ROF model numerical method, the explicit form of the differential equation and implicit format. At the same time the experimental verification of the ROF model edge is well protected,which can effectively remove noise.The ROF model constrainting parameters play a balancing proliferation and inhibition of proliferation of the role of generally taken to be constant. We carry out the improvements proposed the concept and method of calculation of the adaptive parameters, the parameters can automatically adjust its size value according to the local image characteristics, to further strengthen the protective ability of the algorithm for edge.Finally, through a large number of experiments, the wavelet de-noising, the question of how to select the appropriate wavelet basis for image denoising. The ROF model denoising wavelet de-noising effect is not obvious. ROF model parameters are discussed. Denoising and decomposition process of the image, taken on the parameters of the adaptive selection method, and to verify the validity of the method.