平稳小波变换和模糊数学的红外与可见光图像融合  

Image Fusion Method for Infrared and Visible Light Images Based on SWT and Fuzzy Mathematics

作　　者：李婵飞[1] 邓奕[1] LI Chanfei;DENG Yi(Hankou University,Wuhan 430200)

机构地区：[1]汉口学院,武汉430200

出　　处：《计算机与数字工程》2017年第5期874-877,944,共5页Computer & Digital Engineering

摘　　要：提出了一种基于平稳小波变换(SWT)和模糊数学的图像融合算法,该算法具有较好的方向选择性与平移不变性,更适合人类在主观上的视觉认知。先用SWT对原图像进行分解获取低、高频分量,对低频分量采取基于红外、可见光图像不同特性的模糊数学融合准则,高频采用模值取大法融合。实验结果表明,论文算法能更好地改善图像视觉效果,标准差、信息熵和互信息等客观指标都有所提高。A novel fusion method based on stationary wavelet transform and fuzzy mathematics is proposed in this paper.The technique is more suitable for human vision with good directional selectivity and shift-invariance.The original images are decomposed by SWT.According to the different features of the infrared and visible light image,the different membership functions with fuzzy mathematics are utilized as fusion rule for low frequency subband coefficients.And the absolute value is adopted as fusion rule for high frequency subband coefficients.The gray contour target information for infrared image and scene detail information for visible light image are better protected.The experimental results show that the visual effect of fused image is improved efficiently.Besides,the standard deviation,comentropy and mutual information are increased significantly.

关 键 词：图像融合 SWT变换 模糊数学 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]