基于局部能量与模糊逻辑的红外偏振图像融合

针对红外偏振与红外光强图像的成像特征差异,提出了一种基于 NSCT(Nonsubsampled Contourlet Transform)的红外偏振与红外光强图像的新融合算法.该算法首先采用NSCT对源图像进行多尺度、多方向分解,然后对低频子带图像采用局部能量融合规则,融合后再对低频融合图像进行直方图均衡化处理,而对高频方向子带图像采用了模糊逻辑,根据各图像的特征差异来分别选择基于像素、基于区域和加权平均的融合规则,再对得到的各个子带图像进行NSCT重构,从而得到最终的融合图像.通过实验仿真表明本文算法能有效的融合源图像的互补信息,使得该算法在目标识别中具有一定的优势和现实意义.

Infrared Polarization Images Fusion Based on Local Energy and Fuzzy Logic

This paper presents a new fusion algorithm of the infrared polarization and intensity images based on the differences of imaging characteristics between them and NSCT. Firstly, the NSCT is used to perform the multi-scale and multi-direction decomposition of the source images. Then, the low frequency sub-band images are fused by local energy rule and processed by using the histogram equalization, However, the high frequency direction sub-band images are fused by choosing the pixels, the regional and the weighted average fusion rules based on the differences of characteristics in the each image by fuzzy logic. Finally, the various sub-band images fused are reconstructed by the inverse NSCT and get the final fused image. The simulation results indicate that the algorithm is an effective integration of complementary information of the source images and the algorithm has some advantages and practical significance in the target recognition.