基于改进小波阈值函数的医学CT图像降噪

Medical CT Image Denoising Based on Improved Wavelet Transform Threshold Function

针对医学CT图像信噪比低的特点,提出了一种基于软硬阈值的混合函数并结合双树复小波变换的优势对图像进行去噪。首先将图像进行多层分解,然后对分解后的高频子带系数进行新阈值处理,最后通过重构还原图像,经过与其他几种方法的对比,此方法取得了较好的处理效果和峰值信噪比(PSNR)并保留了图像的边缘纹理特征。

According to the characteristics of medical pathology image signal-to-noise ratio is low,this paper proposes a hy？brid based on hard and soft threshold function and combining the advantages of dual tree complex wavelet transform for image denois？ing. First of all,the image is decomposed,then data of each layer undergoes a process with high frequency coefficient threshold,fi？nally,the reconstruction image is restored,Through comparison with several other methods,this method has obtained the good pro-cessing effect and peak signal to noise ratio（PSNR）and preserve the edge of the image texture feature.