模糊逻辑与特征差异驱动的红外偏振图像融合模型

A Fusion Model of Infrared Polarization Images Based on Fuzzy Logic and Feature Difference Driving

利用non-subsampled contourlet transform(NSCT)对红外偏振与红外光强图像进行分解,得到源图像的低频子带和高频方向子带。通过对红外偏振和光强图像差异特征的分析,对低频选取局部能量和局部信息熵提取差异特征,然后利用模糊逻辑融合低频子带的不确定区域,利用特征差异驱动来融合低频子带的确定区域;对高频选取局部边缘信息保留量和局部方差提取差异特征,然后利用模糊逻辑融合高频方向子带的不确定区域,利用特征差异驱动来融合高频方向子带的确定区域。最后利用NSCT对高低频子带进行逆变换得到最后的融合图像。从而建立起基于模糊逻辑与特征差异驱动的红外偏振图像融合模型。实验仿真结果表明,该融合模型可融合源图像互补的差异特征,使其在目标识别和分类中具有一定的应用价值。

Infrared polarization and infrared intensity images are decomposed by non-subsampled contourlet transform （NSCT） to get the low-frequency sub-bands and high-frequency directional sub-bands of source images in this paper. Analyzing the differences between infrared polarization image and intensity image, and using local energy and local information entropy to extract the difference features of the low-frequency sub-bands, we merge the area of uncertainty and the area of certainty of low-frequency sub-bands respectively by fuzzy logic and feature differences driving. Difference features are extracted from high-frequency directional sub-bands by local edge information preservation and local variance, and then the area of uncertainty and the area of certainty of high-frequency directional sub-bands are merged respectively by fuzzy logic and feature difference driving. Finally, by the NSCT inverse transformation of the high and low frequency sub-bands, the final fused image is obtained. Therefore, the fusion model of infrared polarization images based on fuzzy logic and feature difference driving is established. The simulation results show that the fusion model can integrate the complementary difference features and it has certain application value in the target identification and classification.