一种起伏背景下的红外小目标检测算法

A Detection Algorithm of Small Infrared Target under Clutter Background

提出了一种用一维中值滤波进行背景估计的方法。首先分析了红外小目标、噪声及其杂波特性 ,提出了用小窗口的二维中值滤波进行背景估计 ;然后进一步根据红外扫描系统的特点 ,把二维中值滤波简化为一维中值滤波 ,使去背景效果提高 ,并且运算量大幅下降 ;最后对预处理后的图像采用自适应门限进行目标分割。实验表明该算法可以达到较高的检测率 。

Abackground estimation method is presented which is realized by one dimension median filter. Firstly, a background estimation method based on small window and two dimension median filter is presented after detail analysis of the model of small infrared target, noise and clutter. Secondly, two dimension median filter procedure is replaced with one dimension filter according to the property of infrared scanning system, so the effectiveness of removing background is improved and the quantity of operation is far reduced. At last, adaptive threshold is used to segment the targets in the image whose background is removed. Simulation indicates that the presented method has superior effect and satisfies the real time processing of large scale images.