Scale-Aware Pomegranate Yield Prediction Using UAV Imagery and Machine Learning

Monitoring the development of trees and accurately estimating the yield are important to improve orchard management and production. Growers need to estimate the yield of trees at the early stage to make smart decisions for field management. However, methods to predict the yield at the individual tree level are currently not available due to the complexity and variability of each tree. This study aimed to evaluate the performance of an unmanned aerial vehicle (UAV)-based remote sensing system and machine learning (ML) approaches for a tree-level pomegranate yield estimation. Lightweight sensors, such as the multispectral camera, were mounted on the UAV platform to acquire high-resolution images. Eight characteristics were extracted, including the normalized difference vegetation index (NDVI), the green normalized vegetation index (GNDVI), the RedEdge normalized difference vegetation index (NDVIre), RedEdge triangulated vegetation index (RTVIcore), individual tree canopy size, the modified triangular vegetation index (MTVI2), the chlorophyll index-green (CIg), and the chlorophyll index-rededge (CIre). First, direct correlations were made and the correlation coefficientCorrelation coefficient (R \({ }^2\) ) was determined between these vegetation indices and tree yield. Then, machine learning approaches were applied with the extracted features to predict the yield at the individual tree level. The results showed that the decision tree classifier had the best prediction performance, with an accuracy of 85%. The study demonstrated the potential of using UAV-based remote sensing methods, coupled with ML algorithms, for estimating the pomegranate yield. Predicting the yield at the individual tree level will enable stakeholders to manage the orchard on different scales, thus improving field management efficiency.