How Artificial Intelligence Is Enhancing Video Surveillance


For many years, video surveillance and CCTV system have been used to identify and catch thieves, robbers, terrorists and criminals. But if you looked at a video just a few years ago, you would be astonished at its lower quality. Now a day, not only the quality of videos has been improved but also the software behind it. One of the most revolutionary innovations of this millennium is Artificial Intelligence.

Artificial intelligence can now pick you out in a crowd and then track your every move. Japanese firm Hitachi’s new imaging system logs on to a hundred types of qualities and traits of an individual such as mannerisms, facial expressions, thought processes and more besides the physical aspects. Until now we need a lot of security guards and people to review security camera footage. Instead, the new AI software would do the same thing for you. The system can help spot a suspicious individual or find a missing child. An eyewitness could provide a limited description while with the AI software, you can quickly scan its database for a match. In Japan, the demand for such technology is increasing because of the Tokyo 2020 Olympics.

Some AI surveillance software also produce a heat map of where people are massing together. This sort of data can be used to prevent overcrowding. You can also track individuals if you are looking for a suspect.

Similarly, AI surveillance cameras in China watch each and every move of the commuters. As a vehicle passes through them, they can detect its type as well as detect a bicyclist or motorcyclist. The cameras can even recover their trajectory. The cameras are made through an embedded system that put AI technology inside the camera to help the camera to recognise face and objects and stream the original video into a structured data that includes real-time crowd analysis and even mood analysis. China is also testing convenience stores, where registered customers enter with facial recognition and buy items using their phones. Similarly, Ningxia has become a model example – where residents enter compounds with face recognition and track buses on their phones before leaving home. Transport can then be optimised based on user data. They can take masses of user data and also other state data, public & private company data and figure out ways to make cities more livable and secure under the banner of security and convenience. China’s government now has data on almost all of its tech-savvy citizens and technology, for the most part, has made cities better.

According to a report issued by Gizmodo, Singapore government under its smart nation project is also planning to launch a test program of making every lamppost throughout the country a wireless sensor network with facial recognition cameras. The pilot program shall cover almost 110,000 lampposts in Singapore. These cameras would then be used to perform crowd analysis and management, avoiding terrorist activities and investigating crimes. Though it apparently is considered as a security measure, some opponents think the system as an all-seeing eye to target common public, journalist, politicians and protestors.

Leave a comment

Your email address will not be published. Required fields are marked *