Journal of Tropical Oceanography ›› 2012, Vol. 31 ›› Issue (6): 128-134.doi: 10.11978/j.issn.1009-5470.2012.06.020cstr: 32234.14.j.issn.1009-5470.2012.06.020

• Marine Remote Sensing • Previous Articles    

Remote sensing classification of mangrove based on SPOT5 image

SU Xiu1, ZHAO Dong-zhi1, WANG Xiang1,2, ZHONG Cai-rong3   

  1. 1. National Marine Environmental Monitoring Center, Dalian 116023, China; 2. Dalian Maritime University, Dalian 116023, China; 3. Hainan Dongzhaigang National Mangrove Reserve Board, Haikou 571129, China
  • Received:2013-02-06 Revised:2012-05-22 Online:2012-12-01 Published:2013-02-06

Abstract: Taking the Dongzhaigang National Mangrove Reserve in Hainan Island as example, we executed field measurements to obtain mangrove spectral data. On the basis of maximum likelihood classification purification mangrove information, combining with different kinds of mangrove spectral features and vegetation index difference, selected separability threshold, we established decision-making tree classification rule of SPOT5 images based on images between classifications, and tested its classification accuracy. Classification result shows that accuracy for each mangrove type reached 80% and the total accuracy added up to 90%. This research offers a foundation for operational remote sensing monitoring of mangrove ecosystem.

Key words: SPOT5 satellite image, mangrove, spectral feature, remote sensing classification

CLC Number: 

  • TP75