Application of UAV-Based multispectral images for accessing oil palm trees health using online AI Platform

Abstract

The oil palms are a major product in Cambodia to export local and worldwide which provides job opportunities and develops infrastructure over there. In this study, two plots of oil palms have collected by Unmanned Aerial Vehicle (UAV). The Artificial Intelligence (AI) online platform has been used to extracting oil palm trees from very-resolution imagery. In extracting oil palm tree is applied Picterra platform that it has to bring AI-powered for object detection and geospatial analysis. In this way, oil palm trees are counting automatically and manually, the result shows that high overall accuracy by automatically. The oil palm tree health is estimated by using vegetation indices that it has such as Green Normalized Different Red Edge (NDRE), Normalized Different Vegetation Index (GNDVI), and Normalized Different Vegetation Index (NDVI). Vegetation indices maps have been created from UAV-based multispectral images. It has estimated oil palm tree health using Green, Red, Red Edge, and Near-Infrared bands from the multispectral camera. The vegetation indices have been compared with maximum, minimum, mean, and standard deviation in vegetation and chlorophyll content. The NDVI indices are better than NDRE and GNDVI. The spectral profile has been created in both plots based on the pixel value of vegetation indices of reflectance and absorption.