基于线性光谱混合模型的植被丰度遥感信息提取—以梅江流域为例自然地理专业论文.docxVIP

摘要 摘 要 I I 从 TM 遥感影像提取针叶林、阔叶林和低矮植被的丰度信息，对森林资源调 查和生态建设具有重要意义。利用线性混合模型分解得到的 3 类不同植被丰度信 息是遥感软分类方法，通常比遥感硬分类方法更加科学、准确。本文用线性混合 模型从梅江流域 TM 影像中分解得到了针叶林、阔叶林和低矮植被的丰度分布信 息。为改善线性解混的精度，除了经过 FLAASH 大气校正、MNF 变换、PPI 计 算和人-机交互可视化端元选择等一般步骤外，本文尝试增加了地形校正、NDVI 阈值法掩膜处理和反射率归一化处理 3 个环节。 研究结果表明：增加了这 3 个处理环节可以改善线性分解的精度，RMS 分 量图显示均方根误差图的最大值为 0.049，最小值为-0.024，绝大多数的 RMS 值 都小于 0.026，可见总体误差较小。选择野外随机均匀分布的 60 个样本点对模拟 结果进行了线性拟合检验。阔叶林丰度模拟值和实测值之间的相关系数 R2 为 0.8358；针叶林丰度模拟值和实测值之间的相关系数 R2 为 0.8861；低矮植被丰 度模拟值和实测值之间的相关系数 R2 为 0.8202。 关键词：线性混合模型；混合像元；植被丰度；遥感信息；端元 Abstract Abstract II II HYPERLINK /abstraction/ Abstraction of the abundance HYPERLINK /information/ information of broadleaf forest, coniferous forest and scrubby vegetable from TM remote sensing HYPERLINK /image/ images has great importance to HYPERLINK /forest/ forest resource survey and HYPERLINK /ecology/ ecology HYPERLINK /construction/ construction.Extracting the abundance HYPERLINK /information/ information of three different types of HYPERLINK /forest/ forest with linear mixed model is a remote sensing soft classification HYPERLINK /method/ method and the method is HYPERLINK /more/ more HYPERLINK /scientific/ scientific HYPERLINK /accurate/ accurate than remote sensing hard classification. The abundance HYPERLINK /information/ information of broadleaf forest, coniferous forest and scrubby vegetable from TM remote sensing HYPERLINK /image/ images of Meijiang HYPERLINK /basin Basin had been extracted with linear mixed model.To improve the accuracy of extraction of the vegetation abundance HYPERLINK /information/ information with linear mixed model ,in this theses, not only some general methods that include Minimum Noise Fraction(MNF),Pixel Purity Index (PPI) and n-D Visualizer had been adopted, but also three other data processing links that include topographic correction, masking some components using suitable NDVI threshold value and reflectivity normalization had been added. The results showed that three other data process