Detection of Real Time Anomalous Activities using YOLOv7

India has seen an increase in crime rates over the past several years. The National Crime Records Bureau reports that the number of instances has significantly dimbed over the last few years, rising from 383.5 in 2018 to 385.5 in 2019 to 383.8 in 2020. In the present, more crimes are being reported, including accidents, reckless driving, chain stealing, etc. Governments and police departments are taking a variety of actions to reduce these crimes, including enforcing fines, launching awareness campaigns, etc. They are also utilizing modern technologies, such as video analytics through CCTV camera surveillance. To this day, these technologies are solely used to identify reverse driving, no paths, loitering, crowd gathering, and zone monitoring. In order to lessen the aforementioned behaviors, we'll implement a video analytic technique in remaining domains like accidents, rash driving, chain-snatching, etc. ML framework of You Only Look Once version 7 (YOLOv7) for video processing and a control room alarm notification system will be used to develop this video analytic technique. The proposed work intends to establish a secure society in order to benefit the younger, more promising generations.