Implementación de un dron anticolisiones capaz de seguir una trayectoria pre establecida para la elaboración de un mapa de la tasa de ocupación vegetal en una zona.

Abstract

The objective of this degree work was to implement an unmanned flight system with the purpose of providing assistance in specific aspects of calculating and estimating vegetation occupation in a specific area. The development of the prototype was based on three stages. The first stage: to establish, to acquire and to implement the necessary elements; being the Intel Aero development kit, the same as a pre-armed and ready-to-fly drone, an anti-collision system was also made using an Arduino and ultrasonic sensor for obstacle detection. The second stage consisted of the drone calibration and flight image acquisition, for which the QGround Control program was used, it was used to establish communication with the drone to calibrate, configure and monitor the flight parameters. The images were captured with a sports-type camera, placed in the lower part of the drone with which images were captured that cover the space of the route traced by the drone. Finally, an image reconstruction was carried out with the Image-Stitching module of the Matlab development environment, within this we use a segmentation algorithm which allows us to separate the pixels according to their color obtaining the value of vegetable occupation in the whole land, it means that 139 square meters were vegetation. Achieving this way, from aerial images captured by a drone, determine the places that had less density of plants; recommending crop sectors where more agricultural resources should be applied.