Authors: Hang Li*, Indiana State University, Qihao Weng, Indiana State University
Topics: Land Use and Land Cover Change, Remote Sensing, China
Keywords: Temporal consistency,Impervious area, Landsat, Qinzhou
Session Type: Paper
Start / End Time: 3:20 PM / 5:00 PM
Room: Lafayette, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded
The metropolitan area is the final form for the human settlement. As the population is booming around the world, the size of the urban area is sprawling, which attracted a lot of researches on the process of urban developing and land cover classification. The current research attends to improve classification accuracy with the help of some vegetation indices and new methods are applied like Support Vector Machine (SVM), C4.5 and etc. But the methods above mentioned just focus on the characteristics of one individual image. But the land cover change is a gradual procession which has close relationship between the previous and the post. So this research attends to combine a continued series satellite datum to extract the spatial distribution of the built-up area or impervious area. The classification map for our study area Qingzhou, China will be created via C4.5 method. Then the neighbor images can rectified the error in our original image. And our research summarizes the principle of urbanization for our study area based on the consistency of urbanization and input the rules to correct the error for each classification map. The accuracy of classification for images without combination and with combination will be calculated which try to prove whether taking temporal consistency account into classification is useful or not.