①State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China; ②Department of Information Hunan University of Commerce, Changsha 410205, China
Abstract：The construction of available fusion models, such as that based on color space transform, on statistical, on multi-scale analysis is mostly based on the experience. These models have the problem that the parameters are subjectively chosen and can not be adjusted according to the follow applications of the fusion image. A image fusion framework of remote sensing is proposed based on data assimilation and particle swarm algorithm. Under this framework, weights of different attributes are determined according to their importance degree to the following process and object function is composed of the property weights, then the particle swarm algorithm is used to optimize the object function to get suitable image. A group experiments with the help of the entropy, mean gradient, standard variance, spatial frequency and structure similarity verify effectiveness of the framework.
2009, Vol. 31 
