基于NSST变换的超声图像降噪算法  

Denoising Algorithm of Ultrasonic Image Based on NSST Transform

作　　者：蒲久亮 高小明[1] PU Jiuliang;GAO Xiaoming(School of Computer Science and Technology,Southwest University of Science and Technology,Mianyang 621010,Sichuan,China)

机构地区：[1]西南科技大学计算机科学与技术学院,四川绵阳621010

出　　处：《西南科技大学学报》2022年第1期73-79,共7页Journal of Southwest University of Science and Technology

摘　　要：针对超声图像降噪算法在降噪的同时无法很好地保留图像的细节信息问题,提出了一种改进的基于非下采样剪切波变换(NSST)的超声图像降噪算法。首先通过对数变换将乘性噪声转换成加性噪声,然后使用NSST变换对噪声图像进行多尺度分解得到高频子带和低频子带,低频子带使用同态滤波增强细节信息,高频子带采用改进的阈值处理函数降低噪声,对处理之后的高频部分使用梯度域引导滤波(GDGIF)增强图像的细节信息和边缘信息,最后将逆NSST变换后的图像进行指数变换得到降噪后的图像。实验结果表明,该算法峰值信噪比(PSNR)、结构相似性(SSIM)有一定提高,能在去除噪声的同时尽可能保留图像的细节信息。In view of the problem that the denoising algorithm of ultrasonic image could not well preserve the details of the image while denoising,an improved denoising algorithm based on non-subsampled shearlet transform(NSST)was proposed.Firstly,the multiplicative noise was converted into additive noise by logarithmic transform,and then the noise image was multi-scalely decomposed by using the NSST transform to obtain high-frequency subbands and low-frequency subbands.The low-frequency subband used homomorphic filtering to enhance the detail information,the high-frequency subband used an improved threshold processing function to reduce noise,and gradient domain guided filtering(GDGIF)was used for the processed high-frequency part to enhance the detail information and edge information of the image.Finally,the image after inverse NSST transformation was exponentially transformed to obtain the denoised image.The experimental results show that the peak signal-to-noise ratio(PSNR)and structural similarity(SSIM)of the proposed algorithm are improved to a certain extent,and the detailed information of the image can be retained as much as possible while removing noise.

关 键 词：斑点噪声 超声图像 阈值函数 非下采样剪切波变换 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]