News: GMDA to install facial recognition cameras at 10 busy points-17-03-2021
GMDA to install facial recognition cameras at 10 busy points
The Gurugram Metropolitan Development Authority (GMDA) has decided to install facial recognition cameras at 10 busy points in the city to track criminals and strengthen safety measures, officials familiar with the developments said on Wednesday
“We are in the process of installing facial recognition cameras at 12 locations (two already installed) in the city to track movements of criminals in the city and improve security services. The list of locations was given to us by the Gurugram police department. The locations have been selected based on the movement of people, areas where people with difficult crime record are expected (like court complexes),” said a senior GMDA official, requesting anonymity.
Facial recognition technology uses facial features to match with existing information on databases to identify a person. But, due to various reasons such as lack of proper lighting and varied angle of the image, there are chances of false recognitions also in these systems.
The GMDA’s move comes after two cameras installed at Sheetla Mata Mandir and Bus Stand for a trial were found to be working well. The GMDA’s Smart City division tested the functionality of the cameras before deciding to expand the project.
As part of the testing process, pictures of random employees were uploaded on the system managed by the Integrated Control and Command Centre (ICCC) and the persons were made to pass by the camera, which triggered an alarm at the command centre if there was at least a 60% match in the facial features. In the realtime process, the team at the command centre will inform the nearest police control room (PCR) about wanted persons.
Experts are wary of the system as many government institutions in the country are adopting public facial recognition without legal restraints. The central government is also working to develop a National Automated Facial Recognition System to help match images captured from CCTV cameras with existing databases.
Anushka Jain, associate counsel for transparency and Right to Information at the Internet Freedom Foundation, said, “The issue of facial recognition technology is that it is not 100% accurate, whether in the country or across the world. The testing process for these cameras also cannot be completely relied upon as authorities might be testing them in lab conditions with good lighting where facial features will get easily recognised by the camera, resulting in higher accuracy rates.”
“Another concern is that the cameras will capture everyone who comes under their gaze, casting suspicion on everyone. We also don’t know who will have access to this data. The data, if it falls in the wrong hands, can easily be used for extracting personal vendettas as it tracks locations and movements on a continuous basis,” added Jain.
When asked about the accuracy of the cameras, the senior GMDA official said, “These cameras are being tested at major airports also, so they have a good accuracy rate. But if 60-70% facial features match, the alarm will ring and a team will have to go on the ground to identify the person.”
Usha Kundu, assistant commissioner of police (headquarters) said, “In cases of snatching, robbery or firing, these cameras help us in identifying and nabbing the criminals. The positioning of these cameras is decided based on crime records from different areas and the accumulation of crowd. Traffic is also a criterion because these cameras help us with online challans or cases of road accidents.”
The official cited above said that the GMDA is waiting for approvals from authorities concerned, after which the additional cameras would be installed. A tentative list of places for installation of the cameras includes three metro stations, the railway station, the bus stand and a government hospital.
The Smart City team, at present, is in the process of feeding its system with the data of criminals from the police department.
