Odisha has launched tenders for the implementation of a facial recognition project in Rourkela under the framework of Smart City Projects. According to the tender, the facial recognition system will incorporate IP video CCTV cameras and will be proficient in tracking people and analyzing patterns of demonstrations in crowds, and identifying people captured in multiple live streams at the same time.
The tender also states that the system required initially should be proficient in storing up to 10,000 faces on a biometric watchlist in real-time and alert security staff if any unauthorized person is found in a restricted area. It also specifies that the watchlist should be able to scale a million records at least.
The company shortlisted for supplying the system should make sure that it can easily capture images from varied environmental conditions, which also specifies that artificial intelligence-based video analytics will be installed in 110 cameras at different points across the city. A listing of other important capabilities that a system must have includes an ability to identify and authenticate based on individual facial features, detect a face from a photo, match faces from recorded media, detect a face from stored video of any format, and integrate with other systems in the future such as an automatic fingerprint identification system (AFIS), among other requirements.
