Real-Time Handgun Detection in Surveillance Videos based on Deep Learning Approach
T Pavithra, Rajasekaran Thangaraj, P. Pandiyan, Uma Rani M, Balasubramaniam Vadivelu · 2022 International Conference on Applied Artificial Intelligence and Computing (ICAAIC) · 2022
Closed-circuit television (CCTV) systems are vital to prevent security threats to public safety. Nowadays, the use of weapons causes a big security threat in public places and it creates a lot of violence. The quick detection of handguns in public places is very important to avoid or reduce risks. Some places in the world have a lot of crimes that are committed with handguns, even though guns aren't allowed. Closed-circuit television (CCTV) has been used a lot to keep an eye on these situations; it's time to keep an eye on the images. This task is typically carried out by a human, who is more prone to forgetting them due to fatigue or being distracted by something else. Detecting handguns on their own is very important in catching people who use guns to do bad things. Deep-learning based object detectors can't find handguns of different sizes in an unrestricted area. In this work, deep learning techniques are used for automatic detection of handguns in an unconstrained environment through video surveillance footage.