基于Retinex理论和灰度形态学的红外弱小目标增强方法研究

Faint Targets Detection Based on Retinex Theory and Gray Morphological Filtering Method

针对红外弱小目标检测的难点,提出了一种将形态学滤波和Retinex相结合的检测方法。首先采用灰度形态学滤波来对红外图像进行预处理,然后运用Retinex理论对背景进行有效抑制,使图像的对比度和信噪比得到有效增强。该算法用弱小目标的红外图像进行了仿真验证。试验结果表明:该算法可有效的检测出弱小目标,具有检测速度快,抗干扰能力强,易于工程实现的优点。

To address the challenging problem of infrared faint target detection, a method is proposed by combining the gray morphological filtering and Retinex together. Gray morphology is first used to pre-filter the infrared image to be processed, and then Retinex theory is employed to suppress background clutters. As a result, both the image contrast and sinai-to-noise ratio （SNR） are improved. Extensive experiments on several difficult infrared image sequences show that the proposed method is effective in detecting faint targets and powerful in suppressing clutters. Moreover, this method rtms very fast and thus it is promising for practical applications.