Above-Ground Biomass Prediction in Agroforestry Areas Using Machine Learning and Multispectral Drone Imagery

Remote sensing technology applied to agroforestry areas has emerged as an efficient tool to speed up the data acquisition process in decision-making. Unmanned aerial systems (UASs) equipped with multiple sensors are one of the most innovative technologies for predicting Above Ground Biomass (AGB) of trees in agroforestry environment. This study aimed to assess the performance of various vegetation indices, including the Normalized Difference Vegetation Index (NDVI), Normalized Difference Red Edge (NDRE), Green Normalized Difference Vegetation Index (GNDVI), and Optimized Soil Adjusted Vegetation Index (OSAVI), and leaf chlorophyll index (LCI) in predicting biomass accumulation in woody trees. The images used in this study were collected by using multispectral drone images and AGB computed from ground measurements data, which are used as a dependent variable to train our model. The approach uses multiple linear regression and several Machine Learning methods like support vector regression, decision tree regression, and random forest regression analysis. Results are promising, even if further analysis is needed to improve estimation accuracy over an expanded sample area.