摘要
以TM为信息源,对TM多光谱遥感影像的辐射校正、几何校正、研究区实地样地的标定、区划,以及识别模型的提取验证,分析荒漠林树种在TM影像上的机理,分析荒漠林树种光谱特征,建立基于光谱知识的荒漠林树种信息提取的识别模型。结果表明:TM图像的红柳与梭梭的光谱特征明显的差异主要集中在TM4波段上,红柳在TM4的亮度值明显高于梭梭,即红柳亮度值具有在TM4最大的特征;经过对模型的验证,模型判对率为70%。
Based on TM images,the technology of radiation correction,geometric correction,calibration and zoning about the study area field sample,identification model and extraction test were used for TM images processing in this study.Analysis the mechanism and forest spectral features of tree species in the desert from TM images,based on spectral features of tree species in the desert for information extraction and identification model was built.The result showed that their spectral features had distinct differences in TM4 band;the brightness of Tamarix spp.was significantly higher than Haloxylon ammodendron in the TM4 band,the maximum brightness characteristic of Tamarix spp.was in the TM4 band;after the test of the model,the accuracy was 70%.
出处
《福建林业科技》
北大核心
2013年第2期64-69,共6页
Journal of Fujian Forestry Science and Technology
基金
石河子大学高层次人才科研启动资金专项(RCZX200617)
关键词
TM影像
荒漠林树种
信息提取
光谱特征
识别模型
TM images
desert forest species
information extraction
spectral feature
model identification