基于交叉累计剩余熵的多光谱图像配准方法

Registration algorithm of multispectral images based on cross cumulative residual entropy

针对传统互信息图像配准容易产生局部极值的问题,提出将双边滤波器和交叉累计剩余熵结合作为匹配算法,进行多光谱图像的配准。在这种配准算法中,首先针对多光谱图像特点,提出基于概率密度的双边滤波器边缘提取方法,其次采用交叉累计剩余熵代替互信息作为测度函数将参考图像与待匹配图像的边缘进行匹配。双边滤波器的特性是去噪保边,而累计剩余熵比香农熵更具一般性,且该函数可以有效地避免局部极值,去除噪声。实验证明,该方法鲁棒性好,配准效果明显。

In order to solve the problem that classical mutual information images registration may lead tolocal extremum, a new matching algorithm combining the bilateral filter and cross accumulated residualentropy combination was proposed in multispectral image registration. In this algorithm, firstly, accordingto multispectral images characteristics, bilateral filter edge extraction algorithm was put forward based onthe probability density. Secondly, cross cumulative residual entropy （CCRE） was used as the similaritymeasure to match the reference images and transformed images effectively. Bilateral filter is an edge-preserving and noise reducing smoothing filter, and CCRE is more general than Shannon Entropy. Thisfunction can effectively avoid the emergence of the local extremum, overcome noise influence on the thelocal extremum. Experimental results proved that the registration had good robustness, the effect wasobvious.