Chris Williams, Director of Wavestore believes that the demand for facial recognition software will significantly increase over the next twelve months, and particularly so for high security or mission critical applications where a high performance detection capability is required.
“Facial recognition software has already earnt the trust of numerous end-users around the world and there are now clear signs that businesses and organisations in the UK are wishing to take advantage of this exciting technology which can differentiate between welcomed visitors and those who may be intent on endangering the safety of people, property and assets,” said Chris William’s of Wavestore, a company which has fully integrated its video management software (VMS) with the FaceFirst biometric facial recognition solution.
The FaceFirst facial recognition software offers border security, police and military, as well as personnel responsible for security in business environments such as commercial, retail or casinos with the ability to identify suspects, known criminals, terrorists, illegal immigrants, shoplifters and VIPs within seconds of them entering a specific area, allowing appropriate action to be taken.
FaceFirst has just finished installing what is believed to be the world’s largest live face recognition surveillance system at the Tocumen International Airport in Panama. The airport has deployed 155 IP cameras throughout the airport looking for wanted individuals registered in Federal Agency databases. The cameras are running at 30 frames per second and being processed at a speed of a million facial comparisons per second.
“Historically, facial recognition systems were designed to be interactive, in other words one operator with one camera,” said Joseph Rosenkrantz, President and CEO of FaceFirst Biometrics. “This means that somebody walks by a camera which produces possible matches for an operator to review. Well, imagine a system with hundreds of cameras comparing against hundreds of thousands of people on a database. The sheer amount of data would not be humanly digestible. So, we have achieved this large scale by developing our own data processing algorithms which are able to operate
on an infinite scale and distil this deluge of evidence down to humanly comprehensible actionable information. This boils down to millions of comparison requests per second generated by a multitude of sensors, and a system designed to analyze evidence from multiple sources concurrently in order to produce highly accurate identifications of individuals whose faces are on a watch list.”
Once FaceFirst determines that there is ample evidence that a particular person has appeared, it performs instantaneous look ups in multiple national and regional databases and delivers an alert containing the images of the person, the video of the person and all of the associated biographical information via email or other means.
An obvious application for the software is at border crossings. Within just a couple of seconds, whoever needs to know can receive an email containing all the evidence and stats about the person identified, along with the video clip of them passing the camera. The software can also help retailers combat theft by spotting a shoplifter who has been caught in one store only to show up in another of the company’s stores the next week.
“The open architecture of our video management software has enabled us to fully integrate with FaceFirst’s facial recognition solution,” said Chris Williams. “As a result we can offer a combined solution which equips installers to provide their clients with a video surveillance and recording system that can give peace of mind, and can reduce operating costs by making best use of security personnel. It is this kind of technology partnership which I believe will be a key driver in ensuring that facial recognition software will be specified to be part of an increasing number security projects.”