Support vector regression with genetic algorithms for estimating impervious surface and vegetation distributions using ETM+ data

摘要: Accurate estimation of impervious surface and vegetation is a key issue in monitoring urban area assessing urban environments. It has been proved that the nonlinear models for spectral mixture analysis outperform linear in the literature. However, mapping functions require to be predefined which are difficult be determined. Support vector regression (SVR) shown success dealing with problem, such as estimation and prediction. In this paper, genetic algorithm (GA) was employed determine optimal parameters SVR automatically, were applied SVR model. Further, GA-SVR model multi sets (Multi-GA-SVR) was estimate distributions vegetation. The results showed Multi-GA-SVR achieved higher accuracy than single set (Single-GA-SVR) traditional linear mixture (LMM), an overall root mean square error measure (RMSE) 0.15 three distributions. is demonstrated proposed approach promising