Genetic Algorithm-Based Support Vector Classification Method for Multi-spectral Remote Sensing Image

摘要

The traditional classification methods based on asymptotic theory for multi-spectral remote sensing image need the infinite training samples, which is impossible to be satisfied. Support vector classification (SVC) method based on small samples overcome above difficulty. However, the values of hyper-parameters in SVC directly determine the method's performance, which are randomly selected. In order to obtain the optimal parameters, genetic algorithms (GAs) are introduced. Experimental results indicate that this method can not only save time for classification, but also improve the generalization of the SVC model.