基于形态滤波遗传算法对目标红外图像特征检测

Target IR Image Signature Detection Based on Morphology Filtered Genetic Algorithm

将形态滤波和自适应遗传算法结合起来,提出了一个用于目标检测的形态滤波器参数优化遗传算法。通过设计了形态滤波的遗传算法,寻求最佳结构元,以实现对目标的自适应检测。作为2个特殊环境中的应用检测实例,分别对固体推进器尾喷焰红外图像和复杂背景下的弱小目标进行检测。实验结果表明该方法能够更好地适应环境变化,更大程度地提高目标信号检测的信号噪声(SNR)比,有效地检测出复杂背景下的目标。

A morphology filtered parameter optimized genetic algorithm for target detection was put forward, which combines morphology filter and self adaptive genetic algorithm. Self-adaptive target detection was realized through morphology filtered genetic algorithm and seeking of the best framework unit. As two special detecting applications, solid thruster plume IR image and weak small infrared target under complex background were detected. The experimental results testified that this method can adapt to different environment, greatly enhance the signal noise ratio of target detection, and effectively detect target from complex background. Compared with other methods, this method has more practical value.