Denoising Algorithm for CT oral Image Based on Bayesian Threshold and Non-Local Mean

Abstract

Wavelet transform is widely used in speech and image denoising. In the process of oral CT image acquisition, because of the noise caused by imaging principle, human environment and transmission process, it will affect the accuracy of later CT image processing and reconstruction. Therefore, image denoising is an essential part of the CT image preprocessing stage. This paper presents a new Bayesian multi-threshold wavelet transform denoising algorithm. In this method, different thresholds are selected in different generations and directions, and the generated thresholds are used to process the high-frequency coefficients. At the same time, the low-frequency coefficients after wavelet transform are processed with a non-local mean algorithm. Finally, in view of the disadvantages of the traditional soft and hard threshold functions, an improved threshold function is adopted. Compared with bilateral filtering, non-local mean filtering and bilateral filtering algorithms combined with wavelet changes, this method not only improves the peak signal-to-noise ratio and structural similarity, but also makes the image clearer.