The Research Of The Methods Of Image Denoising Based On Wavelet And Contourlet Transform

Reading the image is one of the most main ways to obtain information. As a result, image processing technology is also gradually to the attention of the scholars.In the recent years, Wavelet transform theory is developed rapidly, and its application in image denoising is also gradually increase. However, wavelet transform can't accurately express the image of high-dimensional geometric features. With the rapid development of the related theory of super wavelet, transform gradually prominence and it can retain the image texture better. To a certain extent, it has overcome the weakness of the wavelet. In this paper, the main work is as follows:Firstly, it has studied the relevant principle of the wavelet transform and its implementation algorithm. It has studied the method of image denoising based on the wavelet transform. In the experiment, it has chosen different wavelet bases, and analyzed the wavelet transform in the image to remove the defects existing in the noise. It has studied the related principles and their algorithms of the Contourlet transform, the Wavelet-Contourlet transform and the multi-parameter contraction function- Contourlet transform. Through the Matlab simulation experiment, it makes the comprehensive analysis and comparison of the advantages and disadvantages of the methods above in the field of image denoising processing.Secondly, it has studied the theory of the Cycle Spinning. Because in the process of processing, image signal could produce false Gibbs effect and the visual distortion of image caused by different extent, so it combines the Cycle Spinning theory, and the Contourlet transform, and it has studied the relevant principle and algorithm. It makes the comprehensive analysis and comparison of the advantages and disadvantages of the methods above in the field of image denoising processing.This article combines the Wavelet-Contourlet transform and the Cycle Spinning theory, and puts forward an improved method to remove the noise in the image, andthis method makes up for the Wavelet transform and Contourlet transform lack of translation invariance. Through the Matlab simulation, we can see that the two methods have improved the visual effect of the image denoising by introducing the Cycle Spinning.In the end of the paper, it makes a comparison of the methods above. By comparing the PSNR of the image after denoising algorithm, and the running time of each algorithm, it evaluates the image quality. We can see that the PSNR of the improved algorithm is lower than PSNR of the multiple parameter contraction function- Contourlet transform method, but it greatly shortens the running time and improves the operation efficiency.