Weapons Detection of Criminal Activities Based on Computer Vision
Sakib-Ahmod, Shoyaib Mahmud, Shammir Hossain, Yeasin Arafat, Jakia Rawnak Jahan, Ohidujjaman Ohidujjaman, Raihana Zannat · Advances · 2021
In contemporary the term 'computer vision' is applicable vastly for significant research arena. Many researchers have a keen interest in this field. In this computer vision-based system, there have managed to detect weapons of two types (different guns and knives) to inform the concerned authority of property about crimes ongoing inside the place. While a property is under such criminal activity that the criminals carry guns or/and knives with them, the system detects those weapons through a camera integrated module and runs the necessary functions accordingly. Due to witnessing weapons, the system immediately informs the concerned authority to alert them about the upcoming risk to the property via mobile message and call. Consequently, the proprietor is alerted about the forthcoming risk and takes possible measures to handle the situation. Apart from that, the system can make sound alarms around the place of crime so that the people near to the property can come forward to help. This research is applicable in various properties such as banks, offices, homes or anywhere to protect the valuable properties from being theft or robbed. In this research computer vision and machine learning approaches are applied for the optimal result prediction.