Drone Autônomo com Visão Computacional para Identificação de Pessoas em Áreas Restritas

Abstract

This paper reports the development and analysis of an autonomous drone system designed to detect, recognize, and track individuals in restricted environments using classical Computer Vision methods. The proposed solution integrates the DJI Tello drone with Haar Cascade–based face detection and LBPH facial recognition, both implemented in Python and executed externally. The system was evaluated through a comprehensive set of experiments involving parameter tuning, lighting variation tests, distance-based performance assessment, and measurement of the drone’s responsiveness during autonomous movement. The objective of this study is to examine whether these traditional algorithms, when combined with the hardware constraints of an entry-level drone, can effectively support real-time facial identification and tracking.