Real-Time, YOLO-Based Intelligent Surveillance and Monitoring System Using Jetson TX2

Abstract

AbstractIn recent years, real-time surveillance and monitoring system is getting importance because of security reasons. Government organizations, residential areas, commercial complexes, schools and colleges, industries, borders, and others require a dedicated surveillance system. The traditional surveillance systems do not provide real-time object identification and alerts. This paper aims to design and develop a real-time object detection and alert system (ODAS). The proposed system could be used as a surveillance and monitoring system. The object detection system uses the YOLO framework to detect objects within images and live videos. Object identity and time stamp, along with bounding box image, are marked by the detection system and stored locally. Also, the collected information is transmitted to the server in real time. At the server, the graphical user interface (GUI) application continuously gathers information from different nodes, analyzes it, and fires an alert message if the anomalies/targeted object is found. GUI also facilitates us to analyze the obtained information of an individual node or in the combination that infers the direction of movement made by a detected object and its current position. ODAS also takes the services of the real-time database to provide real-time updates of anomalies on the android phone/tablet.KeywordsYOLOObject detectionSurveillance systemJetson TX2IoT