Medical Image Denoising Algorithm Based On Wavelet Thresholding And Nonlinear Filtering

The medical image denoising technology is the basis of medical image processing and analysis.Its purpose is to eliminate the noise in the image,to restore the original image information to the greatest extent,and to provide reference for clinical medical diagnosis and pathological research.However,due to different imaging mechanisms,equipment and instrument accuracy errors,improper operation and other practical factors,the medical image is easily disturbed by noise during the collection process.At the same time when the image resolution is improved,the image quality is greatly reduced by noise.some tissue boundaries may even become blurred,and the subtle structures are difficult to identify.Therefore,denoising is very important for medical image processing.According to the type of noise and the degree of contamination in medical images,this paper proposes a denoising algorithm based on wavelet threshold and nonlinear filtering technology,which is specifically described from three parts,so as to achieve more targeted noise reduction.(1)Aiming at the X-ray,CT and MRI with single Gaussian noise,an improved wavelet threshold denoising algorithm with good frequency-domain characteristics is selected.The wavelet coefficients after wavelet decomposition are quantized by the threshold in the high frequency direction,the final wavelet reconstruction,to remove Gaussian noise.The simulation results show that the proposed method can effectively remove Gaussian noise,at the same time,and preserve image edge information.(2)For the mixed noise image containing Gaussian noise and impulsive noise,a denoising algorithm based on DT-CWT and adaptive bilateral filtering is proposed.The double-tree complex wavelet of six decomposition directions are more stable with decomposition.After decomposition,the high frequency coefficient is quantized by the threshold,low-frequency coefficients are filtered by bilateral filtering and finally all coefficients are filtered by bilateral filtering to denoise after reconstruction.Experimental results show that this method has a good removal effect on mixed noise,especially the case of light pollution.(3)In order to better solve the problem of heavy noise in medical image,this paper also presents a median filter denoising algorithm based on DWT and correction.Firstly,it is decomposed into four sub-bands(LL,HL,LH,HH)by wavelet transform,where LL is low frequency and LH,HL and HH are horizontal,vertical and diagonal high frequency sub-bands respectively.Then,the improved wavelet threshold is used to process the high frequency coefficients,the low frequency coefficients are retained,the median of the three high-frequency sub-bands after the wavelet threshold processing is modified,and the mixed noise is filtered through the threshold segmentation.Finally,the denoised image is obtained through wavelet reconstruction.The advantage of this method is the combination of the advantages of median filtering and wavelet,the use of coupling coefficient processing makes denoising better,avoiding a simple combination of two methods.In short,the denoising process of medical images not only effectively removes noise,but also preserves the complete boundary and detail information.All of these provide reliable guarantees for later-stage high-level processing.Taking B-placenta images as an example,this paper shows the interference of noise on the accuracy of computer vision recognition,including the phenomenon of feature grabbing and ROI segmentation in regions of interest,specially recognition of contour information of image edge.