Multiclass SVM based land cover classification with multisource data

Support vector machines (SVM) are characteristic of processing complex data and high accuracy. The combination of remote sensing and geographic ancillary data is believed to offer improved accuracy in land cover classification. In this paper, multiclass SVM is introduced to research on land cover classification with multisource data. The classifications of the study site with various methods are given. Experimental results show that SVM have good generation ability on land cover classification. The classification with combination of remote sensing and geographic ancillary data outperforms single remote sensing data in terms of accuracy. Multisource land cover classification based on SVM could gain higher classification accuracy.