Kolkata police are planning to use artificial intelligence-powered software to scan CCTV images in order to identify helmetless bikers and illegal parking in real-time and prosecute offenders. In the city currently, there are over 2,500 CCTV cameras, 125 of which can read licence plates.
Kolkata Police requested a detailed presentation from technology experts in the coming month to help them better understand advanced machine learning tools that are increasingly being used to detect helmetless bikers and illegal parking.
Once the city police have narrowed their focus on the project, it will ask for a detailed presentation. In 2020, 7.3 lakh traffic violations were detected via CCTV footage, out of 39.4 lakh recorded on Kolkata itself.
Not only private companies but there are several academic institutions that are also working on projects to identify helmetless bikers and their licence plates in real-time in the state. Technology is being explored with a view to reducing the risk to officers who attempt to physically stop these motorcycles at their peril. Additionally, police want software models to detect illegal parking and track vehicle licence plate locations.
The software, which is heavily reliant on video data, will be installed in number plate-reading road cameras and on the servers of the Lalbazar police control room's central control room. This can be accomplished by embedding a card into CCTV cameras, which assists in detecting helmetless bikers by sending out an alert. Additionally, when cameras detect vehicles in no-parking zones, alerts will be sent out via artificial image recognition and processing solutions. This application can be implemented in real-time with the assistance of a webcam or a CCTV.
IIT Hyderabad has developed software that uses artificial intelligence to detect bikers without helmets in surveillance footage. The institute recently signed an agreement with the Hyderabad City Police to gain access to the video.
To detect illegal parking, the system utilises an IP camera to continuously capture and transfer image sequences to a base station, where software is used to perform automatic image feature extraction.
According to reports, the proposed software is resistant to the effects of environmental variations, including shadow effects, illumination changes, and varying weather conditions. Both systems are capable of issuing challans. It will be fully automated, with a web interface for monitoring and verifying alerts. It will then be linked to the existing RTO website in order to generate challans and send SMS notifications to riders.
