Drone-Based Sensing for Leaf Area Index Estimation of Citrus Canopy

Leaf Area Index (LAI) is an important parameter in the measuring of crop health. Temporal changes in the LAI provide important information about changes in the structure of the canopy and biomass over time. In this study, RGB images of the top of the canopy are collected by using a drone and through image processing; the coverage of green canopy is calculated from the images. Subsequently, by using the gap fraction, the LAI is estimated through the Beer-Lambert law. The data is collected from Warud taluka of Amravati district of Maharashtra, India. The area is severely under biotic and abiotic stresses. A multi-rotor quadcopter, which can carry a camera, is used to fly over the citrus farm on a predefined path. A camera that is mounted on the drone takes RGB images of the top of the canopy at a continuous interval with 70% frontal and 50% side overlap. These images are stitched together and an orthomosaic image layer is formed. Mathematical models are used to find the LAI from the images. Ground truth data is collected by a ceptometer within two hours of the flight of the drone. The two LAI datasets (LAI from the digital image and the LAI values from the LAI meter) are correlated, with R2 equal to 0.73.