An Improved Wiener Filter Based on Adaptive SNR MRI Image Denoising Algorithm

Abstract

Aiming at the disadvantages of traditional Wiener filtering, a new adaptive noise ratio wiener filtering method is proposed in this paper. The method can identify the noise type according to its histogram distribution type, calculate the mean and variance of noise, and construct the corresponding point spread function.At the same time, the image denoising algorithm based on the improved Wiener filter is realized by estimating the adaptive SNR of the image. Especially for the medical images with different background and foreground, the denoising algorithm proposed in this paper has remarkable effect. The experimental results show that the adaptive SNR wiener filter can achieve better results than the traditional wiener filter by combining the main visual effect and objective PSNR value (the larger the PSNR is the better). The algorithm in this paper can directly find the optimal signal-to-noise ratio of wiener filters, which solves the problem that traditional Wiener filters need to estimate the signal-to-noise ratio continuously.