Based on a spectral characteristic analysis of typical wetland plants, the authors used the object-oriented classification method to extract such specific plants in Hanshiqiao area of Beijing as Phragmites australis, Echinochloa crusgallii and Nymphaea tetragona on the basis of a lot of survey work. First, the authors collected spectral data of these three typical wetland plants in the field, which constituted the basis of spectral correlation analysis and could help make classification between different species. The correlation analysis results of highly distinct spectral band combination and vegetation indexes with remote sensing images were involved in segmentation weight assignment. Second, based on the distribution of typical plants, the authors decided the split-scale of object-oriented (Phragmites australis split-scale being 50, Echinochloa crusgallii split-scale 20, and Nymphaea tetragona split-scale 100). A comparison of different classification methods shows that the classification accuracy of the object-oriented method based on spectral features is about 96%, and the classification accuracy is 87.3% in case it is not based on spectral characteristics. Traditional supervised classification accuracy is only 82.3%. The results show that the object-oriented classification method with spectral feature analysis is effective in the information extraction of Hanshiqiao wetlands typical plants, and the differentiation between spectral bands as well as band combinations will play a key role in that it can highly improve the accuracy of classification.
龙娟, 宫兆宁, 郭逍宇, 赵文吉. 基于光谱特征的湿地湿生植物信息提取研究[J]. 国土资源遥感, 2010, 22(3): 125-129.
LONG Juan, GONG Zhao-Ning, GUO Xiao-Yu, ZHAO Wen-Ji. Information Extraction of Wetland Aquatic Vegetation Based on Spectral Characteristics. REMOTE SENSING FOR LAND & RESOURCES, 2010, 22(3): 125-129.
[1]Li X B,Shi P J. Research on Regulation of NDVI Change of Chinese Primary Vegetation Types Based on NOAA /AVHRR Data[J]. Acta Botanica Sinica,1999, 41 (3):314-324.
[4]Coops H,Hanganu J,Tudor M. Classification of Danube Delta Lakes Based on Aquatic Vegetation and Turbidity. Hydrobiologia[J]. Netherlands,1999,415:187-191.
[5]Dogan O K,Akyurek Z,Beklioglu M. Identification and Mapping of Submerged Plants in a Shallow Lake Using QuickBird Satellite Data[J]. Environmental Management,2008:1-6.
[6]James D,Hurd,Daniel C. Tidal Wetland and Classification from Llandsat Imagery Using an Integrated Pixel Based and Object Based Classification Approach[C]// ASPRS 2006 Annual Conference.Reno,2006:1-5.
[10]Wilson E H,Sader S A. Detection of Forest Harvest Type Using Multiple Dates of Landsat TM Imagery[J]. Remote Sensing of Environment,2002,80:385-396.