Design and Implementation of Vehicle Monitoring System: a Case Study of Bouesti.

ABSTRACT
Unauthorized vehicle parking posed significant challenges in institutional environments, such as blocking access paths, damaging lawns, and constituting inefficient use of parking facilities. Traditional methods like manual patrolling and stationary cameras are labor-intensive, time-consuming, and failed to provide real-time detection. This project addressed these challenges by developing a real-time Unauthorized Vehicle Parking Detection System that leverages You Only Look Once (YOLOv11) for vehicle detection and DeepSORT for tracking. The system uses captured video as input, detects vehicles using YOLOv11, and tracks their movements across frames using DeepSORT to identify unauthorized parking based on predefined zone mappings of unauthorized areas. The system then generated notifications mails to the assigned personnel in charge of the operations to ensure prompt actions. Testing in a controlled environment showed the system achieved accurate detection, with reliable tracking and an average processing speed of 28 frames per second. The system effectively identified unauthorized cars using video inputs. This study highlighted the potential of combining YOLOv11 and DeepSORT for scalable and automated parking management. By leveraging captured video, the system provided a cost-effective and environmentally friendly solution that aligned with Sustainable Development Goals by reducing unauthorized parking and protecting green lawns. Future enhancements, including better lighting support, automated zone updates, and real-world deployment, are recommended to optimize the system further.