摘要
基于Landsat8遥感数据源采用方差分析的方法进行各波段数据的可分性度量研究,从而筛选出高温目标识别的最佳波段。通过对不同地表覆盖类型样本光谱进行分析,提出多种光谱波段的组合方式并对其进行可分性度量分析,得出研究区高温目标识别的最优光谱指数形式。通过地类样本数据,检验高温目标的识别精度为91.3%,达到了较好的识别效果。研究表明:采用Landsagt8遥感数据构建高温目标光谱指数可以实现地表高温目标的有效识别。
The variance analysis method is used to study the separability of each band of the Landsat8 remote sensing data source and the best band of high temperature target recognition is selected. In the study area,the combination of different spectral bands and the separability analysis are put forward to obtain the optimal spectral exponent form of high temperature target recognition based on the analysis of the spectra of different types of surface coverage. The recognition accuracy of high temperature target is 91. 3% verified by the land sample data. The results show that the spectral index of high temperature target constructed from Landsagt8 remote sensing data can effectively identify high temperature target.
出处
《世界地质》
CAS
2017年第4期1297-1302,共6页
World Geology
基金
高等学校博士学科点专项科研基金新教师类资助课题(20110061120067)资助
关键词
遥感
高温目标
方差分析
可分性度量
光谱指数
remote sensing
high temperature target
variance analysis
separability analysis
spectral index
