基于局部信息熵及其分布特性的图像融合算法

Image Fusion Algorithm Based on Local Information Entropy and Its Distribution Property

提出了一种新的基于HIS模型的图像融合算法,根据源图像在有向梯度域中的局部信息熵及其分布情况确定融合准则。对实验结果的主观定性评价和客观定量分析说明,新算法在有效保持空间信息的同时,产生的光谱畸变较小,对全色图像中的噪声有一定的免疫力,从而克服了现有同类算法光谱失真严重和对源图像中噪声敏感的缺点。

This paper proposes a new multispectral and panchromatic image fusion algorithm based on HIS model. The new algorithm determines fusion rule according to local information entropy and its distribution property in orientated gradient domain. The subjective qualitative evaluation and objective quantitative analysis of experiment results prove that this new algorithm can produce fusion image that preserves spatial characteristics effectively and introduces less spectral distortion. It is also less sensitive to the noise in source panchromatic image compared with some existing algorithms.