Image Denoising Using WT Based On GA And SVM

The real-life images are always corrupted by noise. When the noise is strong, it can affect image segmentation, recognition and comprehension. Traditional denoising methods can filter noise, but at the same time they make the image details fuzzy. Recently, with the improvement of wavelet theory, wavelet analysis has penetrated into many fields. Meanwhile, wavelet applied to image denoising successfully, and many new image denoising algorithms based on wavelet have been proposed. But few people synthetically apply Genetic Algorithm(GA) and wavelet transform(WT)N Support Vector Machine(SVM) and wavelet transform to image denoising.Based on serious study of using the threshold shrink procedure to denoise, the paper suggests two denoising methods according to wavelet transform: GA&WT method and SVM&WT method. In the previous method, firstly classical genetic algorithm is improved based on pinciple of gene reconfiguration. Reverse and random logical cross operation are defined and used to implement gene reconfiguration. Experimental results show that the IGA has much higher convergence speed and stability than classical GA and extends the results in the paper[41]. Then based on improved GA, every scale best thresholds are solved after decomposed by mu(?)tiscale wavelet transform. Reconstructed signal is obtained by using the inverse wavelet transform after all coefficients are dealt with soft threshold method. Experimental results show that the method is effective and can obtain optimal signal-to-noise ration.In the SVM&WT method, modulus maxima denoising method is studied firstly. Then SVM are trained by modulus maxima and then classes the wavelet coefficients. Reconstructed signal is obtained by using the inverse wavelet transform for not noised signal coefficients. Experimental results show that the SVM&WT method is easy to be realized and can obtain optimal signal-to-noise ration.