An object-based classification approach for surface water detection

G. Xiao, D. Tien

Research output: Book chapter/Published conference paperConference paperpeer-review

Abstract

Detection of areas of water on the land surface via aerial imagery is highly desirable for assisting land management. As near-infrared (NIR) energy tends to be absorbed by water, this property of low spectral reflection is usually utilized in analyzing water resources on land. However, the spectral reflection of shallow water varies significantly. It is difficult to distinguish such areas from the background by traditional land cover classifications. This paper proposes an object-based classification approach for automatically detecting water areas from aerial imagery with red, green, blue, and NIR bands. To overcome the problem of inadequate class definition in conventional region-based classifications, the water areas (we focus here on farm dams) are divided into a number of classes, and we newly introducing a specific feature of spectral differences and proposed a decision tree (DT) approach to select the features required for each class. In this way, we separate small water areas from other objects with significant reduction in false alarms caused by the diversity of spectra in the feature space. Experiments show that the proposed approach has good capability to distinguish shallow water areas from other objects in wetlands.

Original languageEnglish
Title of host publication6th International Conference on Information Technology and Applications, ICITA 2009
Pages10-14
Number of pages5
Publication statusPublished - 2009
Event6th International Conference on Information Technology and Applications, ICITA 2009 - Hanoi, Viet Nam
Duration: 09 Nov 200912 Nov 2009
http://www.icita.org/previous/icita2009/home.htm (Conference website)

Publication series

Name6th International Conference on Information Technology and Applications, ICITA 2009

Conference

Conference6th International Conference on Information Technology and Applications, ICITA 2009
Country/TerritoryViet Nam
CityHanoi
Period09/11/0912/11/09
OtherThe 6th International Conference on Information Technology and Applications (ICITA 2009) will be held in Hanoi, Vietnam on 9 - 12 November 2009. The conference is an annual event and is recognized as one of the major series of conferences in Information Technology (IT in short). ICITA continues to retain its tradition of high quality conference and will open up an opportunity for academics and industrial professionals worldwide to exchange their knowledge of the state-of-the-art research and development in information technology and applications, and to establish and extend professional networks. This conference addresses the importance that IT professionals stretch across narrowly defined subject areas and constantly acquire a global technical and social perspective. ICITA 2009 has offered such an opportunity to facilitate cross-disciplinary and social gatherings.
Internet address

Fingerprint

Dive into the research topics of 'An object-based classification approach for surface water detection'. Together they form a unique fingerprint.

Cite this