摘要
在小波变换理论的基础上,提出了一种结合小波分解和脉冲耦合神经网络(PCNN)的遥感图像融合新方法.首先对两幅已经配准的原始遥感图像进行小波多尺度分解,得到低频子带系数和各带通子带系数;其次对低频子带系数采取一种基于边缘的方法以得到融合图像的低频子带系数;对各带通子带系数提出了一种改进的基于PCNN的图像融合方法来确定融合图像的各带通子带系数;最后通过逆小波变换重构图像得到融合后的图像.仿真结果和评价指标结果表明,此方法更好地保留了原图像中的有用信息,提高了融合图像的质量.
This article proposed a new fusion algorithm of Remote Sensing image based on wavelet decomposition and pulse coupled neural networks（PCNN）.First,obtains the low frequency coefficient and various bandpasses coefficient though wavelet decomposition of two primitive matched Remote Sensing image.Then,obtains the low frequency innertube coefficient of image by the edge method.Next,obtains the various bandpasses innertube coefficient of image by PCNN.Finally,obtains the restructuring fusion image after the counter wavelet decomposition.The fusion image indicated that this algorithm not only retaining the original information well,but only improving the fusion image quality.
出处
《华中师范大学学报（自然科学版）》
CAS
CSCD
北大核心
2012年第1期117-120,共4页
Journal of Central China Normal University：Natural Sciences
基金
国家自然科学基金项目(6080303361062003)
