Vegetation Canopy Coverage Estimation Using Physical Models

Abstract

Based on the physical models of PROSPECT, SAIL and porosity model, hyperspectral data and canopy coverage data of different combined scenes were simulated. According to the simulated data, we chose four sensitive bands and four sensitive vegetation indexes highly correlated to vegetation canopy coverage, and analyzed the correlation between sensitive bands, sensitive vegetation indexes and canopy coverage. Then we built a regression model of canopy coverage with EVI highly correlated with canopy coverage. At last, we verified this model by experimental data from ground measurement experiment. It shows that there is a high correlation between EVI and canopy coverage and the regression model built by EVI can produce an effective result and the RMSE is less than 0.09.