30 October 2009 Study on the classification of urban wetlands based on RS and GIS
Author Affiliations +
Proceedings Volume 7498, MIPPR 2009: Remote Sensing and GIS Data Processing and Other Applications; 74982Q (2009) https://doi.org/10.1117/12.833963
Event: Sixth International Symposium on Multispectral Image Processing and Pattern Recognition, 2009, Yichang, China
Abstract
aIn this paper, band-combining operation is applied to extract water body area of three different period remote sensing pictures. Based on the Beijing wetland management classification system of wetlands and wetland patches census data, through the superposition of space under the counter-analysis and recursive speech, build a VBA classification, different stages of the wetlands water are separated automatically in order for the organic connection and unity of data from wetland administration section and remote sensing data. The results show that: (1) Decision tree is a good way to get the whole water information. All the information of water can be extracted by tm2 + tm3> tm4 + tm5, but there are still some mixed information. Towns and clouds can be removed when tm5 and tm7 less than a specific threshold, and shadow of the mountain can also be removed well when tm3-tm4 are greater than a specific threshold. (2) The urban wetlands can be obtained rapidly with the secondary development of VBA functions in GIS. (3) This classification method has high accuracy, the overall classification accuracy is 91.8%, kappa statistics is 0.88, and this method avoid post-processing work, is very time-sensitive.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Weiguo Jiang, Jing Liu, Li Liu, Shaoxue Sheng, Peng Hou, "Study on the classification of urban wetlands based on RS and GIS", Proc. SPIE 7498, MIPPR 2009: Remote Sensing and GIS Data Processing and Other Applications, 74982Q (30 October 2009); doi: 10.1117/12.833963; https://doi.org/10.1117/12.833963
PROCEEDINGS
8 PAGES


SHARE
Back to Top