基于LWT和递归最小类内绝对差的红外小目标检测  

Infrared Small Target Detection Based on LWT and Recursive Minimum Within-cluster Absolute Difference

作　　者：纪守新[1] 吴一全[1] 

机构地区：[1]南京航空航天大学信息科学与技术学院,江苏南京210016

出　　处：《信号处理》2010年第10期1484-1488,共5页Journal of Signal Processing

基　　金：国家自然科学基金资助项目(60872065)

摘　　要：针对存在背景干扰和噪声情况下的红外弱小目标检测问题,提出一种基于提升小波变换(LWT)和递归最小类内绝对差的检测方法。一方面先利用提升小波对原始图像进行去噪,再利用Top-hat算子抑制背景;另一方面先利用Top-hat算子抑制原始图像的背景,经提升小波去噪后,再进一步使用Top-hat算子;上述两方面得到的图像求和即为预处理图像。然后采用递归最小类内绝对差阈值选取方法分割预处理图像。针对红外小目标图像进行了大量实验,并与基于形态滤波及基于小波和形态学的红外小目标检测方法进行了比较。结果表明本文方法提高了信噪比,检测率分别提高15%和10%。Aiming at the detection problem of dim target in infrared image that contains background interference and noise,a detection method is proposed based on lifting wavelet transform and recursive minimum within-cluster absolute difference.Firstly,the image is preprocessed.The original image is denoised based on lifting wavelet transform,then the background of denoised image is suppressed by Top-hat operator.At the same time,the background of original image is suppressed by Top-hat operator,then Top-hat operator is further used after the residual image is denoised.Addition of the above-mentioned two resultant images gives the preprocessed image. Secondly,the preprocessed image is segmented using the threshold selected by recursive minimum within-cluster absolute difference. Lots of experiments are done with infrared images including small targets,and a comparison is made with the detection methods for infrared small target based on morphological filter and based on wavelet and morphology.The results show that the signal-to-noise ratio of the suggested method is improved,and detection rate increases by 15%and 10%,respectively.

关 键 词：红外弱小目标检测 提升小波变换 TOP-HAT算子 递归最小类内绝对差 

分 类 号：TN911.73[电子电信—通信与信息系统]