The need to identify and locate suspects fast can be met by various smart search analytics software today, which sifts through overwhelming amounts of video data based on certain search criteria and returns results with lightning speed.
With criminal activities on the rise and threats of terrorist activities more imminent than ever, video surveillance has become an indispensable tool in pinpointing and identifying suspects. The Boston Marathon bombers, for example, were caught by surveillance cameras in the area, and law enforcement officials were able to use the footage to eventually locate the killers.
Yet with piles upon piles of video footage, which can easily amount to terabytes of data, searching for a particular individual can be a daunting task. “After a major incident, investigations can be extremely resource consuming,” said Erez Goldstein, Senior Product Marketing Manager at Qognify. “Trawling through large amounts of video footage to isolate the right person out of the crowd and track their route is an extremely complex task to conduct manually. It takes a huge amount of manpower, and it can be slow, which is far from ideal in a time pressured scenario where you need to learn fast – Who was involved? Who were they in contact with? How they arrived? Where they might have been before acting their crime?”
The need to identify and locate suspects fast can be met by various smart search analytics software today, which sifts through overwhelming amounts of video data based on certain search criteria and returns results with lightning speed. One example is Qognify’s Suspect Search, which addresses two key areas in which smart search has a crucial role to play – real-time incident management and post event investigation.
“New search tools enable security professionals to work in real-time, or near real-time, rapidly narrowing down data from 100 or 1000s of camera feeds in a matter of minutes, in order to locate a suspect, missing person or object,” Goldstein said. “In a post-event scenario, the ability to automate much of the process saves huge amounts of time and resource, improving the likelihood of a successful resolution.”
With various smart search analytics in the market, Goldstein said Qognify makes its solution unique with the following:
Searching based on a human “signature”: Qognify has developed the ability to identify a specific individual and search for them in the video footage, based on unique features of that individual, not just his or her face.
Being camera agnostic: Aside from very basic image quality and size requirements, the solution is completely independent of the camera from which the image is obtained. The entire processing and algorithmic handling is done on the server.
Color variation compensation: In real life surveillance, cameras have color variation due to location and illumination, among others. The Suspect Search algorithm knows how to compensate on color variation between cameras to bring accurate results across many types and location of cameras.
Besides law enforcement, the product has applications in other areas, too. “There is quite a wide usage pattern across many vertical markets,” Goldstein said. “These include transportation (air and land), city centers, critical facilities (for example corporate facilities and hospitals, and large public events.”