Research On Image Denoising Algorithm Based On Wavelet Transform

The wavelet transform has been applied in the field of image denoising for many years, but with the development and improvement of wavelet theory and the new wavelet proposed, the area of its application in image denoising is constantly developing. Wavelet threshold denoising method, which has simpler and less computation, has been recognized and valued. So wavelet threshold denoising method has been widely used.The main job of this paper is to do some research on wavelet threshold denoising method and gain some appropriate improvements. In this paper, a new threshold function is proposed based on soft and hard threshold function. The threshold function overcomes the disadvantage of the hard threshold function which is not continuous, well preserves the local features and image edges, and effectively suppresses ringing and other visual distortion of the image. Simulation results show that the improved threshold method can eliminate noise more effectively, compared with the traditional soft and hard threshold function.In the process of being polluted, image is not only polluted by a single noise frequently, but by a variety of mixed noise. There are impulsive noise and Gaussian white noise in the main interference of image. Therefore, this article combines the median filter and the new wavelet threshold method. We first perform the median filter in spatial domain to obtain a pre-denoised image, then play the new wavelet threshold method in wavelet domain. The method can inherit the advantage of the median filter and wavelet denoising method and effectively reduce the interference of mixed noise. Finally, We use MATLABL7.10 to play some simulations. Results show that the improved denoising method has superiority and the ability to reduce the mixed noise.