Smart face-detecting cameras to scour public places in Kerala soon

Mail This Article
Thiruvananthapuram: Police have to keep up with smart criminals who use latest gizmos and technology to carry out their nefarious activities. The Kerala Police is all set to install cameras with Artificial Intelligence (AI) that can track down criminals and locate missing people even among crowds in real time.
To begin with AI cameras would be installed at the bus stations at Thampanoor and East Fort (Kizhakkekotta) in the Kerala capital, Thiruvananthapuram. The first control room making use of the facility would be opened at the Cantonment police station in Thiruvananthapuram.
The Kerala Development and Innovation Strategic Council (K-DISC) has invited tenders from start-ups in order to execute the project.
The last date for receiving tenders is September 24.
The Motor Vehicles Department is also planning to install AI cameras to nab those violating traffic rules. KELTRON has already made a project proposal in this regard.
How the AI system works
AI cameras are cameras that use AI programmes to efficiently deal with collected images and videos.
The photos of criminals in the police records would be uploaded and if any of them are caught by the camera, the information would be swiftly passed on to the police control room. The system would use the facial recognition feature to identify the criminals.
The photos of missing people would also be uploaded into the system.
The software would be developed to recognise several faces at the same time. The cameras would be placed at a height of 7 ft to identify the face. The information passed on to the control room would include the percentage of similarity with the photo on the police records. The video footage, including the suspect people, would also be sent to the mobile phones of officers concerned.
The police had earlier installed face-detection cameras at the Sannidhanam at the famed Sabarimala temple to identify trouble-makers last year when agitations were raging over women's entry at the hill shrine. Around 200 such people were then identified with these cameras. However, this was not entirely powered by artificial intelligence.
Apart from real-time monitoring, the software would also be developed to identify people from recorded video footage and from group photos.