期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2007
卷号：XXXVI-4/W54
页码：179-184
出版社：Copernicus Publications
摘要：We present a new object-oriented land cover classification method based on Support Vector Machine (SVM) by fusing spectral and textural information of High-Resolution (HR) aerial imagery and Digital Surface Model (DSM) information of Light Detection and Ranging (LIDAR) data in the same urban area. It synthesizes the advantages of digital image processing (image fusion, image segmentation), Geographical Information System (GIS) (vector-based feature selection) and Data Mining (intelligent SVM classification) to interpret image from pixels to segments and then to thematic information. Compared with the pixel-based SVM classification where the multi-source imageries were used, the results showed that the proposed object-oriented method by fusing of multi-source imageries could share redundant and complementary information, and also provided greater classification detail and accuracy, especially, various kinds of shadows were correctly identified, and the computational performance for classification were improved. Moreover, the land cover classification map could update GIS database in a quick and convenient way
关键词：Aerial imagery; Light Detection and Ranging (LIDAR); Image Fusion; Object-oriented Classification; Land-cover ; Classification; Support Vector Machine (SVM)
