基于高光谱数据的植被覆盖度估算研究

Study on Vegetation Coverage Estimation Based on Hyperspectral Data

在研究区域内选择了120个样方,根据草地类型的不同光谱特征构建相应模型并予以识别。基于高光谱影像不同通道的遥感影像值,选择常用的高光谱遥感植被指数作为评价指标,将各种不同的遥感指标与实测的草地覆盖度进行回归分析,找出相关性最高的遥感指数对整个研究区域的植被覆盖度进行估算。研究结果表明:高光谱的NDVI指数与实测数据相关性最高,与实测数据拟合程度高达95%,用于雅鲁藏布江的植被覆盖度具有很好的可行性。

In this paper,120 quadrats were selected in the study area,and the corresponding models were constructed and identified according to the different spectral characteristics of grassland types.Based on the remote sensing image values of different channels affected by hyperspectral,the commonly used hyperspectral remote sensing vegetation index is selected as the evaluation index,and various remote sensing indexes are analyzed and regressed with the measured grassland coverage,so as to find out the remote sensing index with the highest correlation and estimate the vegetation coverage of the whole study area.The results show that the hyperspectral NDVI has the highest correlation with the measured data,and the fitting degree with the measured data is as high as 95%.It is feasible to be used for the vegetation coverage of Yarlung Zangbo River.