基于CHRIS数据的HDVI植被指数的植被覆盖度估算研究

HDVI CHRIS Data for Vegetation Index Estimation of Vegetation Coverage Based on Research

选取江西省余干县的一景CHRIS/PROBA影像5个观测角度(±55°、0°和±36°)的反射率,在CHRIS数据处理后得到的NDVI基础上得到HDVI,并与野外实测的草地、灌木、针叶林、针阔林和阔叶林的VFC建立相关性分析,以及对于模型拟合度的决定系数R2的影响因素进行分析。结果表明,在±55°下,针阔林的多项式拟合度最高,植被类型模型的拟合度均值也最高。其次是灌木拟合度较高。最差的是针叶林指数模型,植被类型的整体拟合度最不理想。在±36°下,针阔林多项式拟合度最高,植被类型的整体拟合度也最高,其次是阔叶林拟合度较高。最差的是针叶林的指数模型,植被类型的整体拟合度也是最不理想的。

This paper selects a scene CHRIS/PROBA images of Yugan county Jiangxi province with 5 observation angle （ ±55° ,0°and ±36° ） refleetivity. HDVI can be obtained on the basis of NDVI, NDVI is obtained after CHRIS data processing. And the correla-tion analysis was established with the VFC of the field measured grassland, shrub, coniferous forest, coniferous and broad leavedmixed forest and broad leaved forest. As well as the influence on the fitting degree of the model decision coefficient R2 factor analysis.Results show that in ±55°, the highest degree of polynomial fitting in the coniferous forest, the average fitting degree of vegetationtype model is also the highest. Followed by a higher degree of fitting of shrubs. The worst is the coniferous forest index model and thewhole fitting degree of vegetation type is also the least. In ±36°, the highest degree of polynomial fitting in the coniferous forest, theaverage fitting degree of vegetation type model is also the highest. Followed by a higher degree of the broad leaved forest. The worst isthe index model of coniferous forest, its whole fitting degree of vegetation type is also the least.