Land use change detection based on remote sensing classification and pixel comparison--a case study
In this paper,land use changes were studied based on remote sensing classification and comparison of overlapping pixels methods in Jiading district of Shanghai from 1989 to 2006.Multi-source data including four epochs of representative TM images (in years of 1989,1995,2001 and 2006) and the vector topographic map were used in our study.Land use classes were first extracted by remote sensing classification after the image preprocessing such as geometric correction and registration,and a change detection method by comparing the pixels in overlapping images was then presented.Based on the classification results of land use classes with four epochs of TM images,the changes of land uses in three time intervals that 1989-1995,1995-2001 and 2001-2006 were therefore detected by the detection method,and the conversion process of land uses classes for nearly 20 years was further analyzed in the study area.The conclusion was finally made that the continuous construction of building,road and greenbelt in Jiading district in past decades costs the expenses on a lot of cropland,forest and orchard land.
Remote sensing classification Change detection Pixel comparison Land use
Xue Zhang Xiaohua Tong Miaolong Liu
Department of Surveying and Geo-informatics,Tongji University,No.1239 Siping Road,Shanghai 200092,China
国际会议
第16届国际地理信息科学与技术大会(16th International Conference on GeoInformatics and the Joint Conference)
广州
英文
2008-06-28（万方平台首次上网日期，不代表论文的发表时间）