基于小波分析的双色红外弱小目标检测算法  

A Dual-Band Infrared Dim Target Detection Algorithm Based on Wavelet Analysis

作　　者：史晓刚[1] 白晓东[2] 李丽娟[1] 徐振亚[1] 韩宇萌[1] 刘佳鑫[2] 

机构地区：[1]中国空空导弹研究院,河南洛阳471009 [2]中国人民解放军第5715工厂,河南洛阳471009

出　　处：《红外技术》2016年第8期688-692,共5页Infrared Technology

基　　金：航空科学基金项目(20110112007)

摘　　要：针对单波段红外弱小目标检测难度大、信息量少的问题,提出一种基于小波分析的双色红外弱小目标检测算法。首先运用小波滤波器对双色图像进行一级分解,然后提出一种软、硬阈值折衷法对高频小波系数进行处理,低于阈值的系数将被直接归零以抑制背景;通过采用一定的策略对双波段高频图像进行融合并将低频系数归零后,运用小波反变换得到滤波图像,最终采用多帧累积检测完成弱小目标的检测。实验证明本文算法能有效提升目标信噪比,具有良好的弱小目标检测能力和实时性。It is very difficult to detect infrared dim target in single-band because the comparative lack of message. Aimed at this problem, an algorithm of infrared dim target detection in dual-band based on wavelet analysis is introduced. The dual-band images are decomposed by wavelet filter firstly, and a compromise model which is between hard-threshold and soft-threshold models is introduced to dispose the high frequency wavelet coefficients, and the coefficients which are lower than the threshold will be changed to zero in order to restrain the background. The dual-band high frequency images will be fused by use of certain strategies and the low frequency images will be changed to zero, so the filtering image will be obtained by wavelet inverse transform, and the dim target will be detected by several frames accumulative detection finally. The experiments show that this algorithm can improve the target＇s signal-noise ratio effectively, and it has favorable abilities in infrared dim target detection and real-time.

关 键 词：小波分析 双色 弱小目标检测 信噪比 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]