基于改进LOG算子的雷达图像边缘检测算法  

Radar image edge detection algorithm based on improved LOG operator

作　　者：李平 张勇[1] 田忠彬 吕西昆 王晴晴[1] LI Ping;ZHANG Yong;TIAN Zhongbin;LYU Xikun;WANG Qingqing(Air Force Early Warning Academy,Wuhan 430019,China;Joint Logistics Support Force,Wuhan 430019,China)

机构地区：[1]空军预警学院,武汉430019 [2]联勤保障部队,武汉430019

出　　处：《空天预警研究学报》2024年第1期16-20,共5页JOURNAL OF AIR & SPACE EARLY WARNING RESEARCH

摘　　要：雷达图像中固定地物杂波的边缘轮廓对运动目标检测有重要作用.针对传统的高斯-拉普拉斯(LOG)边缘检测算法在对雷达图像进行边缘检测时对噪声敏感,易影响图像边缘轮廓信息的提取问题,提出了一种改进的LOG边缘检测算法.首先采用改进的均值滤波和双边滤波对图像进行平滑去噪;然后用拉普拉斯算子计算二阶方向导数,计算零交叉点得到图像的边缘位置信息,从而获得连续、完整的地物边缘轮廓;最后对雷达图像采用原始LOG和本文改进的LOG算法进行了对比实验.实验结果表明,在强地物杂波环境下,与原始LOG算法相比,改进的LOG算法提高了边缘的检测精度,改善了图像边缘连续性,从而可提高雷达目标的检测概率.The edge contour of fixed ground clutter in radar images plays an important role in moving target detection.In order to solve the problem that the traditional LOG edge detection algorithm is sensitive to noise in radar image edge detection,which is easy to affect the extraction of image edge contour information,an improved LOG edge detection algorithm is proposed.Firstly,the improved mean filtering and bilateral filtering is used to smooth and denoise the image.Then Laplacian operator is used to calculate the second order directional derivative,and the zero crossing point is calculated to obtain the edge position information of the image,so as to obtain the continuous and complete edge contour of the ground object.Finally,comparative experiments are conducted on the radar images processed by the original LOG and the improved LOG algorithm.The experimental results show that,in the strong clutter environment,compared with the original LOG algorithm,the improved LOG algorithm can improve the edge detection accuracy and the continuity of the image edge,thus able to increase the detection probability of the radar targets.

关 键 词：雷达图像 高斯-拉普拉斯算子 边缘检测 均值滤波 双边滤波 

分 类 号：TN958[电子电信—信号与信息处理]