Uso de sistemas de aeronaves remotamente pilotadas (RPAS) en el monitoreo de plantaciones forestales

Abstract

This research was carried out at the facilities of the La María experimental farm of the Quevedo State Technical University, with the objective of identifying the feasibility of using PGR in the monitoring of forest plantations, diagnosing the current state of PGR management in forestry activities and defining the characteristics of the tree vegetation cover of the aforementioned farm. The vegetation area under study corresponds to approximately 17 hectares, located at coordinates: 666781.916E - 9879909.569N 17M, which maintains a lagoon separated into three sections, with a total of 1,918 hectares, two marshes separated by rural roads that make up 1,156 hectares. The overflights were carried out with the RPAs: Mavic Pro, Mavic 2 Pro and Mavic Air 2 and the processing of the images was done with the mobile application DroneDeploy, the same that was used to design the flight plan, because this application apart from creating the flight plan, also allows creating the orthophoto (union of all the photos recorded by the drone). The YOLO object detector was used, considered fast and accurate, which makes it ideal for machine vision applications, allowing the creation of object recognition systems in real time.