结合Lee滤波的NL-Means声呐图像滤波方法  

NL-Means sonar image filtering algorithm combined with Lee filtering

作　　者：田原嫄[1] 雷玉峰 郭海涛 TIAN Yuanyuan;LEI Yufeng;GUO Haitao(School of Mechanical Engineering,Northeast Electric Power University,Jilin 132012,China;School of Marine Science and Technology,Hainan Institute of Tropical Oceanography,Sanya 572022,China)

机构地区：[1]东北电力大学机械工程学院,吉林吉林132012 [2]海南热带海洋学院海洋科学技术学院,海南三亚572022

出　　处：《河南科技学院学报（自然科学版）》2024年第5期46-52,共7页Journal of Henan Institute of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金(61661038);海南省自然科学基金(420CXTD439);中国三亚市科技创新专项(2022KJCX83)。

摘　　要：声呐图像中存在的散斑噪声不仅极大地影响了图像质量,还对图像的后续分割、增强、边缘检测等增加了很多负面影响.针对这一问题,提出了结合Lee滤波的NL-Means算法:先用Lee滤波算法对声呐图像进行一次滤波,再用滤波后的图像计算NL-Means算法中的权重部分的权值,最后用NL-Means算法结合该权重进行滤波.改进的滤波方法避免了权重计算过程中由于噪声的影响而使导致权重偏小的问题,精确了NL-Means算法中的权重分配.实验证明,改进方法对声呐图像的散斑噪声不仅有更好的抑制效果,还提升了图像的质量.Scatter noise is prominent in sonar images,and the presence of scatter noise greatly affects the quality of the image,adding a lot of negative impact on the subsequent segmentation,enhancement,edge detection and other processing of the image.To address this problem,an NL-Means algorithm incorporating Lee filtering is proposed,which first uses Lee filtering algorithm to filter the sonar image once,and then uses the filtered image to calculate the weights of the weights in the NL-Means algorithm,and then finally combines this weight with the NL-Means algorithm to perform filtering.This method avoids the problem of small weights due to the influence of noise in the weight calculation process and accurately assigns the weights in the NL-Means algorithm.Experiments have proved that the method has better suppression effect on the scattering noise of sonar image and improves the quality of the image.

关 键 词：声呐图像滤波 Lee滤波 NL-Means算法 评价指标 

分 类 号：TN911.73[电子电信—通信与信息系统]