The RS Image Classification Based On SVM Method With The Ecosystem Public-spirited Wood's Type-A Case Study Of Yuhang District In Hangzhou

The ecosystem public-spirited wood means the forest focuses on behavior existence, keeping life and society economic stabilizing development, creating better ecosystem environment.Carrying on the monitoring of ecosystem public-spirited wood management is the base work for the realization of the scientific, normal, precise and modern construct of public-spirited wood.This paper deal with the RS image classification based on the SVM method, combine a space characteristic etc. information, carry on a classification to the IKONOS high spatial resolution image, monitor of implement to ecosystem public-spirited wood.Combine to this method and tradition method to carry on more analytical.This shows that the RS image classification based on the SVM method can solve the image classification fragmentation,the low accuracy etc. and have advantage to the studying speed, from orientation ability, expression etc.The aim of this paper is to discuss a method to inquiry into the classification method of one public-spirited wood's typing of kind's ecosystem to the high spatial resolution RS image, providing theories basis and data support for the development which builds up a forestry information network and"numerical forestry".Support Vector Machine(SVM) is a new method for pattern recognition based on the statistical learning theory.It takes structure risk minimization (SRM) as principle.By mapping input data into a high dimensional characteristic space in which an optimal separating hyperplane is built.SVM presents a lot of advantages for resolving the small samples, nonlinear and high dimensional pattern recognition, as well as other machine-learning problems such as function fitting.At the present time applying SVM to pattern recognition area is set as a focus in the study of SVM.In the beginning SVM is used to solve the two-category classification but the multi-category classification.How to extend it to multi-category classification effectively is still being studied.At present many algorithms that extend SVM to solve multi-category classification are developed which called Multi-category Support Vector Machine (M-SVM).At last it shows that the SVM algorithm has a high precision of classification and a good ability of generalization. And all these indicate that SVM have a goodfuture in remote sensing image recognition.