利用PRO4SAIL与支持向量机回归的组合模型反演植被等效水厚度  

Quantitative Inversion Vegetation Equivalent Water Thickness by Combined Model of PRO4SAIL and the Support Vector Regression

作　　者：李丹娜[1] 郭云开[1] 朱善宽[1] 刘宁[2] 刘磊[2] 蒋明[2] 

机构地区：[1]长沙理工大学交通运输工程学院,湖南长沙410076 [2]长沙理工大学测绘遥感应用技术研究所,湖南长沙410076

出　　处：《测绘通报》2017年第8期50-55,共6页Bulletin of Surveying and Mapping

基　　金：国家自然科学基金(41471421;41671498)

摘　　要：为监测路域植被生态环境,利用遥感影像和辐射传输模型物理基础实现了对植被冠层等效水厚度(EWT)的估测。提出了利用PRO4SAIL与支持向量机回归的组合模型对等效水厚度进行反演的方法。选取Landsat7 ETM+影像,结合实测数据探索验证了PRO4SAIL与支持向量机回归的组合模型的植被参数反演的实用性和准确性。研究表明,该组合模型具有较好的预测能力,反演得到的等效水厚度含量精度较高,为支持向量机模型应用于遥感影像反演植被参数提高了有力支撑。To monitor the ecological environment of the road vegetation, remote sensing image and the radioactive transfer model are used to estimate vegetation equivalent water thickness. In this paper, a combined model of PRO4SAIL and support vector regression is presented to estimate vegetation equivalent water thickness. We explore the feasibility and accuracy of inversion vegetation parameters of the combined model of PRO4SAIL and support vector regression by the Landsat7 ETM＋image and measured data. Research shows that the combined model has good prediction ability, and the accuracy of the equivalent water thickness is high. It also can improve the strong support for support vector machine model applied to remote sensing image inversion of vegetation parameters.

关 键 词：等效水厚度 PRO4SAIL 支持向量机回归 植被参数反演 

分 类 号：P237[天文地球—摄影测量与遥感]