Automated fence surveillance by use of drones
Typ
Examensarbete på kandidatnivå
Bachelor Thesis
Bachelor Thesis
Program
Publicerad
2023
Författare
Berg, Marcus
Cewers Bredberg, Joakim
Falsen Lindqvis, Albin
Johansson, Johannes
Kronqvist, Carl
Olsson, Malte
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Abstract
The widespread availability and affordability of UAVs (Unmanned Aerial Vehicles) have opened up new possibilities for their utilization, enabling the development of novel applications. This paper discusses how to use a commercial UAV from the manufacturer DJI to autonomously collect footage of a perimeter fence. The collection phase first involves careful planning of a trajectory. Furthermore, it requires advanced software for controlling the vehicle. This footage is then analyzed using a neural network or a line detection algorithm to detect abnormalities such as fallen trees or holes. This research concludes that data collection using a drone is feasible and is a comparably uncomplicated problem to solve. However, the analysis step is more involved and will require more work to be fully operational. The project shows potential for providing value for facilities that requires a high level of fence integrity.