基于形态小波变换的红外图像目标检测方法

Infrared Image Targets Detection Based on Wavelet Transform and Mathematical Morphology

文中提出了一种基于形态小波变换的海空背景下受强杂波、噪声污染的红外图像目标检测算法。算法利用了小波变换的多分辨率、多尺度特性,将要检测的图像分解到不同频率的多个尺度上,再采用形态学的背景估计和形态滤波技术,对分解后的子图像进行处理、小波重构。仿真实验表明,该算法可较强的抑制云层、海浪以及海天线的强杂波背景和强噪声的干扰,可检测出信杂比(SCR)为2的目标,适用于舰载红外警戒系统。

The algorithm based on wavelet transform and mathematical morphology is presented to resolve the infrared image detection polluted heavy yawp. First by multiresolution and pyramid decomposition, the components with different spatial resolutions and different directions can be acquired. The component images are filtered with mathematical morphology, then be structured with wavelet athwart transform. The experimenal results show that the algorithm can eliminate the strong clutter of clouds, ocean waves, sea level and strong noise. And the algorithm is especially suited for surveillance system which can detect the infrared target with signal to clutter ratio equal to 2.