Computer Laboratory Surveillance System: Robbery Scene Detection and Alerting
Chanawee Sae-Ung, Pawika Nadeethae, Akara Prayote, Porawat Visutsak · 2022
This paper presents the smart surveillance system using YoLo v3 to detect the robbery scene and send an alert notification through Line application. Our system is capable of detecting the moving objects (CPUs and monitors) and people in computer laboratory. The surveillance system can recognize the authorized persons and captures the suspect who try to steal laboratory's device (system assumption: an unauthorized person is carrying CPU and a monitor out of the laboratory). We used YoLo v3 for CPUs and monitors detection, face detection and recognition, and intersection over union (IOU) technique for the bounding box. The sensitivity, specificity, and accuracy of our system are 0.72, 0.64, and 0.68, respectively.