An Improved Image Denoising Method Based on Multi-scale Correlation in Wavelet Domain

In this paper, an improved image denoising method based on multi-scale correlation is proposed. The cumulative histogram of multi-scale correlation in wavelet domain is calculated, and the threshold is determined by means of the cumulative histogram. Wavelet coefficients at each scale are shrinked using a new soft thresholding function. Experiments show that this method can remove the noise while preserve the significant details