[Application of ANFIS in in-situ measured hyperspectral data for vegetation chlorophyll content estimation]

Hyperspectral reflectance and green degree of Platanus orientalis L. and Populus tomentosa Carr. leaves were measured by the ASD portable spectrometer and the portable chlorophyll meter SPAD-502, respectively. The chlorophyll concentration retrieval models based on 10 common vegetation indexes were established, and the ANN-BP model which used wave bands with larger correlation coefficient as input variables was established for chlorophyll content estimation. Finally, the ANFIS model was established to inverse vegetation chlorophyll content using hyperspectral data. The results showed that normalized difference vegetation index can inverse chlorophyll content better than other vegetation index, and the determination coefficients R2 of models of Platanus orientalis L. and Populus tomentosa Carr. were 0.795 7 and 0.754 6, respectively. The determination coefficients R2 between the predicted and the measured chlorophyll content based on ANN-BP models of Platanus orientalis L. and Populus tomentosa Carr. were 0.935 2 and 0.917 1, respectively. ANFIS model which is a good method to be applied to hyperspectral data for estimation of vegetation chlorophyll concentration can greatly improve vegetation chlorophyll concentration estimation accuracy, and the determination coefficients R2 between the predicted and the measured chlorophyll content of Platanus orientalis L. and Populus tomentosa Carr. were 0.935 2 and 0.917 1, respectively.