Study on classification methods of remote sensing image based on decision tree technology

• Published in: 2011 International Conference on Computer Science and Service System (CSSS) pp. 4058 - 4061
• Main Authors: Wenming Shen, Guozeng Wu, Zhongping Sun, Wencheng Xiong, Zhuo Fu, Rulin Xiao
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2011
• Subjects: Accuracy; Automatic Classification; Data mining; Decision trees; Decision-tree; Land use/Land cover; Pixel; Remote sensing; Remote Sensing Image; Satellites; Spatial Data Mining; Training
• ISBN: 9781424497621; 1424497620
• DOI: 10.1109/CSSS.2011.5972192

In order to improve and enforce environmental monitoring ability, especially in fields of large scale monitoring and dynamic monitoring, the Environmental Satellite will be launched in 2008 in China. Before the Satellite is launched, necessary pre-research work has to be done. Considering future ecological monitoring demand, we have paid more attention to land use/land cover classification method based on the Satellite's CCD sensor. In this article, we compared the decision tree classification technology with other classic automatic classification technologies using Landsat ETM+ image data and GIS data of Tangshan City in Hebei, China. The result of this study showed: accuracy of decision tree classification compared with the classic automatic classification technologies was improved by 18.29%, Kappa coefficient was increased about 0.1878; classification accuracy was improved about 19.52% when DEM and its derivative data were used as ancillary data in the mountainous area, Kappa coefficient was increased about 0.281; the classification accuracy was improved by 15.86% when the DN(Digital Number) values were converted to at-satellite reflectance values; tasseled cap transformation could cause classification accuracy to be reduced appreciably accompanied by compression of data amount.