Land-use Classification of Islands Based on Decision-tree Method
YANG Xi-guang1,2, HUANG Hai-jun1, YAN Li-wen1, DU Ben-xu3
1. Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China; 2. Graduate University of Chinese Academy of Sciences, Beijing 100049, China; 3. Dalian Forestry Bureau, Dalian 116023, China
Abstract With Moye island of Rongcheng city in Shandong province as the study area, the authors investigated the application of decision-tree method to land-use classification based on SPOT 5 satellite data. The results show that the decision-tree method is suitable for classification,with the mean precision and Kappa index being 86.46% and 0.8414 respectively. A comparison with the other traditional classification methods shows that the precision of decision-tree method is obviously higher, suggesting that the method of decision-tree classification has better applicable potential than the other methods in land-use classification research.
XU Hong-gen,WANG Jian-chao,ZHENG Xiong-wei, et al. Land-use Classification of Islands Based on Decision-tree Method[J]. REMOTE SENSING FOR LAND & RESOURCES,
2012, 24(2): 116-120.