Study on Image Denoising Method Based on Multiple Parameter Shrinkage Function

In transform based image denoising methods, how to modify the transform coefficients is an important problem. In wavelet image denoising, two-dimensional tensor product wavelet has isotropy, with poor selectivity, making it difficult to describe the high dimensional geometric features of images. With the development of multi-scale transform, Contourlet transform is emerging prominently. In this study, the advantages of soft threshold and hard threshold shrinkage functions are combined and a multiple parameter shrinkage function (MPSF) is proposed for image denoising. To verify the effectiveness of MPSF, it is used to denoise images polluted by Gaussian white noise. Experimental results show that the proposed shrinkage function is effective, and the denoised images have satisfactory visual effect, with significantly improved image quality metrics such as peak signal to noise ratio.