结合时相和波谱信息的植被信息提取

Extraction of the vegetation information based on temporal and spectrum information

本文利用2002年陕西省ETM卫星影像及相关植被地理信息,在其影像上选取覆盖全省包含不同地物类型的85个地面样地(Ground Truth),通过2003年陕西省MODIS卫星影像获取所选样地全年的NDVI时间序列曲线。采用决策树分类方法,结合NDVI时间序列曲线,实现基于时相和波谱信息的植被分类。最后通过混淆矩阵与Kappa系数等方法对分类结果进行正确率评价,结果表明,文中所给方法优于传统分类方法,所得结果与其他调查结果相一致。

This paper chose 85 Ground Truths（GT） which included different types of features in ETM image of Shaanxi province in 2002 using interrelated geographic information of vegetation.The NDVI time sequence curves of the GT were obtained by MODIS data of Shaanxi province in 2003.Based on temporal and spectrum information,the decision tree classification method combined with NDVI time sequence curves of GT was realized.Finally,confusion matrix and Kappa coefficient were used to evaluate the accuracy of the classification.The result showed that the method was superior to the traditional classification methods,and the results were consistent with other findings.