Genetic Algorithm Based Optimization after Sub-pixel/Pixel Spatial Attraction Model for Sub-pixel Mapping

The spatial resolution of hyper-spectral remote sensing image is not enough to describe the distribution of land cover classes in the mixed pixel, sub-pixel mapping (SPM) is a promising way to predict the location of end-member at the sub-pixel level, based on the fraction images which were generated by spectral un-mixing. In this paper, a novel method was proposed to realize SPM. The proposed method was contained two main steps: sub-pixel/pixel spatial attraction model (SPSAM) is used to generate the initial results, and genetic algorithm (GA) as the post-process method to optimize SPM. The experiment was tested on two sets of data: simple artificial images, synthetic image. The results show the proposed algorithm has the better accuracy than the original SPSAM method.