Maximum Local Energy Method for Multispectral Image Fusion in Remote Sensing System
Maximum Local Energy Method for Multispectral Image Fusion in Remote Sensing System
复制标题
DOI：
--
复制
发表时间：
2012
影响因子：
--
通讯作者：
Lifeng Zhang;Huimin Lu-;S. Serikawa
中科院分区：
文献类型：
--
作者：
Lifeng Zhang;Huimin Lu-;S. Serikawa
提出了一种适用于多光谱遥感图像融合的基于多分辨率分析的图像融合方法。该方法致力于多光谱低分辨率遥感图像与高分辨率全色图像的融合。利用超小波变换提取遥感图像系数的高分辨率区域。BWT是一种多分辨率分析，其基函数是分辨率逐渐增加的方向边缘。BWT的优点是很好地适应固化奇异和点奇异。这种方法有两个明显的优点。首先利用方向细节系数对图像边缘进行匹配，比在可分离小波域中得到的效果更好。其次，低频系数间最大局部能量（MLE）选择策略更为有效。实验证明，极大似然估计是一种新的遥感图像融合方法，具有较好的融合效果。并对小波变换后的融合结果进行了比较。
This paper presents an image fusion method suitable for Multispectral remote sensing image fusion, based on multi-resolution analysis in remote sensing system. This method dedicates to fuse multispectral low-resolution remotely sensed images with a more highly resolved panchromatic image. The high-resolution regions of each remotely sensed image coefficients are extracted by beyond wavelet transform (BWT). BWT is a multi-resolution analysis, whose basis functions are directional edges with progressively increasing resolution. The advantage of BWT is well adapted to cured-singularities and point-singularities. There are two obvious virtues of this method. First directional detail coefficients matching image edges, which is better than obtained in separable wavelet domain. Secondly, selection strategy of maximum local energy (MLE) between low-frequency coefficients is more effective. Experiments have proved that the MLE is a new way for remotely sensed image fusion, which can get well performance. Fusion results in beyond wavelet transform are also compared.