基于小波和Context模型的海面红外弱小目标检测  

Dim Small IR Sea Target Detection Based on Wavelet and Context Model

作　　者：孙国栋[1] 吉书鹏[1] 周桢[1] 

机构地区：[1]中国空空导弹研究院,河南洛阳471009

出　　处：《红外技术》2010年第2期97-100,共4页Infrared Technology

摘　　要：红外图像中弱小目标的检测一直是图像处理领域的热点和难点。介绍了一种基于小波变换和Context模型空间自适应滤波的红外弱小目标检测算法。该方法首先利用小波滤波器抑制大部分背景杂波,然后采用基于Context模型的空间自适应滤波器对高频系数做进一步处理,提高信噪比,最后采用基于Bayes分类算法的迭代门限进行图像分割。实验结果表明,与高通滤波相比,该方法能够更有效地检测出弱小目标。Small target detection from infrared image（IR） has been a hot and difficult issue in image progressing filed. In this paper, a detection method based on wavelet transform and Context model spatial adaptive filter is presented. Firstly, a wavelet filter is used to suppress the background clutter, then a Context-based adaptive filter is utilized in the further processing of the wavelet detail coefficients, which can improve the signal to noise ratio（SNR） greatly and preserve the candidate targets information effectively. Finally, the candidate target is segmented from background using iterative threshold based on Bayes classification algorithm. Experimental results show that it can detect the target more effectively compared with the high-pass filtering.

关 键 词：小波变换 Context模型 红外弱小目标检测 自适应滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]