双树复小波和独立分量分析的红外小目标检测  

Infrared Small Target Detection Based on Dual-tree Complex Wavelet Transform and Independent Component Analysis

作　　者：吴一全[1,2] 纪守新[1] 尹丹艳[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 [2]南京大学计算机软件新技术国家重点实验室,江苏南京210093

出　　处：《兵工学报》2010年第11期1431-1437,共7页Acta Armamentarii

基　　金：国家自然科学基金资助项目(60872065);航空基金资助项目(20105152026);南京大学计算机软件新技术国家重点实验室资助项目(KFKT2010B17)

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,提出一种基于双树复小波变换和独立分量分析(ICA)的检测方法。对图像作预处理:利用双树复小波对原始图像进行去噪,再利用Top-hat算子抑制背景;从原始图像减去通过快速独立分量分析(FastICA)分离出的背景图像,用双树复小波去噪;上述2方面得到的图像求和即为预处理图像。采用模糊Tsallis-Havrda-Charvat熵选取阈值来分割预处理图像。针对红外小目标图像进行了大量实验,并和基于快速独立分量分析的目标检测方法、基于形态滤波的目标检测方法进行了比较。结果表明,本文方法抗噪性强,具有更为优越的检测性能。Aimed at the detection of infrared dim target in background interference and noise,a detection method was proposed on the basis of dual-tree complex wavelet transform and independent component analysis（ ICA）. Firstly,the original image was denoised by using the dual-tree complex wavelet transform,and the background was suppressed by using Top-hat operator. The background image separated from the original image by fast independent component analysis （ FastICA） was subtracted from the original image,and the residual image was denoised by using the dual-tree complex wavelet transform. The sum of the above-mentioned two resultant images gave the preprocessed image. Secondly,the preprocessed image was segmented by using the threshold selected by fuzzy Tsallis-Havrda-Charvat entropy. Lots of experiments were carried out with infrared images including small targets and a comparison was made with the detection method based on the fast independent component analysis and the detection method based on a morphological filter. The experimental results show that the proposed method in this paper is stronger in antinoise performance and more superior in detection performance.

关 键 词：信息处理技术 红外弱小目标检测 双树复小波变换 独立分量分析 模糊Tsallis-Havrda-Charvat熵 

分 类 号：TN911.73[电子电信—通信与信息系统]