Automatic System to Alert Assault Risk to Drivers in Shared Transport Services

The research and development of computer vision-based technologies to mitigate issues in public safety has been a constant challenge. This work focuses on the small prototype development for assault risk detection to drivers on shared transportation service vehicles. The system implements an SSD Mobilenet object detector algorithm, trained for weapons automatic detection such as guns or knives on Nvidia Jetson Nano hardware equipped with a webcam video camera and GSM data transmission module that interconnects it with a web server for the automatic generation and sending alert messages via SMS and email to the registered emergency contact. This system could significantly enhance the safety of shared transportation service drivers by providing a faster and more efficient response from the authorities in potentially critical situations.