Smart evaluation of videos for unusual-event identification in automated vehicle monitoring systems

Abstract

Many methods have been tried to perfect vehicle recognition, which includes quickly locating the same objects in cameras with different sensor types and perhaps propagating trustworthy information across cameras to improve detection. In this study, we look at various important issues and provide a new method for spotting certain abnormalities. We suggest an automated vehicle monitoring system (AVMS) to enhance detection accuracy. If the current reading from a certain local monitor is out of the ordinary, that monitor will issue a notification, and all of those alerts will go toward deciding whether or not an abnormal event has occurred. The effective implementation of any large-scale surveillance system relies on our algorithm meeting several parameters. It simply takes a few minutes to set up and is completely hands-free. The experimental findings demonstrate the effectiveness and simplicity of the suggested approach, with the assembled image quality being on the level with that of state-of-the-art techniques.