Lecture Notes in Computer Science, 2008, Volume 5370/2008, 740-749, DOI: 10.1007/978-3-540-92137-0_81

Multi-Classifier Systems (MCSs) of Remote Sensing Imagery Classification Based on Texture Analysis

Hongfen Li, Guangdao Hu and Jiang-feng Li

View Related Documents

Abstract

This article concerns methods of improving the accuracy of land cover maps using Very High-resolution Satellites (VHRS).It discusses two methods for increasing the accuracy of classifiers used in land cover mapping. One is texture analysis using GLCM method and the other is multiple classifier system (MCSs) using voting rules. A case study of QuickBird Imagery of an area in Chenggong County of Yunnan Province is conducted based on an analysis of QuickBird imagery. The experiment results show that these two methods can improve the accuracy greatly. The applying of texture bands makes an increase of 2.6816%, and the MCSs make an increase of 3.9512%.

Keywords  multiple classifier system (MCSs) - co-occurrence probability - texture analysis - land-cover classification

Fulltext Preview

Image of the first page of the fulltext document