A novel framework to detect suspicious packages using object detection techniques

One of the main tasks of security in public is to make sure no unattended articles are left anywhere, ensuring that no unwanted articles are left where they can harm people. Most of such articles are bags, boxes or other such similar articles. For a place such as a metro where every 15 minutes people depart or arrive an unattended article left behind is inconspicuous and usually not noticed. This paper devices and application to set a timer on such articles using object detection. This timer alerts if an article is in such a fast-moving vicinity for too long so as to alert the guard to fetch the article and check. This ensures not only less lost property but also people's safety. In previous papers mentioned in the literature survey the Object detection is used to di eventuate between different object. In this project is built a customized object detection system which only detects people and belongings such as a bag, backpack and other such carriers using YOLO (You Only Look Once) algorithm in a dark ow environment using python. It then sets a timer on those carriers if the timer exceeds a limit set by the user it sends an alert to the user.