Estimating chlorophyll concentration of paddy using airborne hyperspectral data based on in situ spectral

Biophysical Parameter Data such as chlorophyll are important for agriculture management. Chlorophyll is the main pigment in photosynthetic. Generally healthy plants that can grow at maximum level have more amount of chlorophyll than the unhealthy one. It needs special algorithm to obtain good accuracy to estimate the content of paddy chlorophyll with hyperspectral imaging. The object of this research develops in situ reflectance into algorithm model of content estimation for paddy leaves chlorophyll for airborne hyperspectral imaging. In this research, several vegetation indexes such as normalized difference vegetation index (NDVI), modified simple ratio (MSR), Triangle Vegetation Index (TVI), Soil Adjusted Vegetation Index (SAVI), modified / transformed chlorophyll absorption ratio index (MCARI, TCARI), and (MCARI/OSAVI and TCARI/OSAVI) integration are used to form estimation model using regression method. In addition, model approach with Multiple Linear Regression (MLR), Principle Component Regeression (PCR), and Partial Least Square Regression (PLSR) are also used. From regression process, four ground models which have strong correlation (R2 ≥ 0.5) toward paddy chlorophyll are found. Those four models are MLR 5 band (699.4 nm, 2428.2 nm, 774.8 nm, 1148.8 nm dan 531.2 nm), TCARI/OSAVI (683.9 nm, 714.5 nm, 805.5 nm dan 546.6 nm), Red Edge Absorption Clhorophyll Index (774.8 nm, 714.5 nm dan 638. 3 nm), and PLSR (44 Band) PC no 11. The most optimal model for paddy chlorophyll content estimation is PLSR (44 band) with 0.754 R² and 1.44 SPAD unit of RMSE. Copyright