Imagus Technology Pty Ltd: Taking Facial Recognition to the Next Level

Imagus Technology Pty Ltd: Taking Facial Recognition to the Next Level

Brian Lovell, CTO, Imagus Technology Pty LtdBrian Lovell, CTO
Over a decade ago, when suicide bombers with knapsacks full of explosives attacked Central London, it was one of the worst terrorist iniquities on British soil. The bombing proved that there existed no practical method to process thousands of hours of surveillance video apart from human examination. “Virtually every conventional face recognition approach failed miserably on such low quality images where cooperative face recognition systems are very slow and require expensive camera and lighting installations to capture high quality faces. A key technological advancement would be the ability to be able to reliably identify a person from surveillance video. This is where our expertise comes into the limelight,” explains Brian Lovell, CTO, Imagus Technology Pty Ltd. A provider of core technology for advanced real time systems that reliably detect and identify persons easily, Imagus Technology specializes in non cooperative face recognition from surveillance cameras, mobile phones, social media and wearable cameras.

Along with traditional border and access control applications, the Imagus Facial Recognition Technology opens up a new range of opportunities for unconventional, profitable business solutions based on non-cooperative face recognition. For instance, a system developed by Imagus is used to determine the audience for an electronic billboard in Times Square. The technology not only counts the number of people passing by, it also measures the age and gender of persons looking at the billboard. Additionally, the company also offers a system that measures how long it takes for a customer to be served in a multi queue fast food establishment as well as counting and measuring the customer traffic. “Today, systems are being installed to recognize shoplifters as well as the good customers in major shopping centers. Such self-funded retail business systems can be interconnected and centrally managed to instantly look for persons of interest across wide areas and place their activities on a map,” observes Lovell.

Imagus specializes in recognizing very difficult and low resolution faces at extremely high speed from commodity cameras with natural lighting

Originally developed for counter-terrorism, intelligence and policing applications, over the years Imagus has collaborated with businesses in a range of industries. The company’s work with the banking sector to validate identity documents via a mobile smart phone has proved to be immensely convenient and practical compared to requiring customers to visit the bank. “The heart of the problem is reliably matching a selfie to secure identity documents without human intervention,” Lovell says.

“Our current focus is on cloud-based transcontinental surveillance system for retail and domestic clients. Once the face recognition client systems are installed and configured, the system is completely managed over the public internet via a web browser interface. Each company or a chain of stores gets their own private cloud space to handle their private watch lists and collect their recognition alerts,” explains Lovell. The highly scalable system vastly improves security by alerting the guards and police to potential problems before they happen to create safer cities.

Driven to solve highly challenging problems through excellent scientific research, Imagus has a large biometrics R&D capacity, in addition to a defense grade video coding team. Despite the complexity of face recognition as a technology, the company also provides complete video face recognition systems running on mobile phones and even augmented reality glasses such as the R7 Smartglasses from Osterhout Design Group. The Imagus app can match a face in real time to the one in a database with resolutions as low as of twelve pixels between the eyes. As Imagus’ face recognition technology continues to be embedded in everyday lives via TVs, refrigerators and smart homes, Lovell’s mission for the future is simply to be the very best in the homeland security arena. “Our vision is that we want every computing device to recognize its owner and other persons as a standard service,” concludes the CTO.