CBL - Campus del Baix Llobregat

Projecte llegit

Títol: Sistema de detección, seguimiento y control PID para drones de interior y exterior con modos de juego (2)

Estudiants que han llegit aquest projecte:

Director/a: VALERO GARCÍA, MIGUEL

Departament: DAC

Títol: Sistema de detección, seguimiento y control PID para drones de interior y exterior con modos de juego (2)

Data inici oferta: 16-07-2025     Data finalització oferta: 16-03-2026


Estudis d'assignació del projecte:
    GR ENG SIST AEROESP
Tipus: Individual
 
Lloc de realització: EETAC
 
Paraules clau:
UAS, DEE, PID, Redes Neuronales
 
Descripció del contingut i pla d'activitats:
 
Overview (resum en anglès):
This Bachelor's Thesis aims to develop a software interface for basic drone control and object tracking using computer vision techniques, with a focus on the creation of different interactive game modes. The system has been designed to operate with both an indoor drone (DJI Tello) and an outdoor drone (Hexsoon), adapting the control logic to the characteristics and constraints of each platform.
The adopted methodology is based on the incremental development of the system, progressively validating each implemented module. Color contour detection techniques and neural network models trained with YOLO have been used for object tracking. Drone motion control is performed using a PID controller, which was experimentally tuned through real and simulated tests. In addition, several game and demonstration modes were developed to clearly visualize the system's behavior. For the outdoor drone, both simulation environments and real-world tests were carried out, incorporating different modifications and additional mechanisms.
The obtained results show that the system is capable of stable object tracking in different scenarios, both indoors and outdoors, validating the correct operation of the detection, control, and visualization modules. During testing, several limitations were identified, mainly related to lighting conditions, the computational cost of neural networks, and the limited time available for exhaustive testing, especially when using a Raspberry Pi in real-time. Nevertheless, the project provides a functional and extensible foundation, supported by a GitHub repository that enables its integration into an academic environment and its further development in future work.

© CBLTIC Campus del Baix Llobregat - UPC