A video forensics expert has been talking to SecurityNewsDesk about the use of CCTV footage in the investigation of the Boston marathon bombing.
The analyst, who was commenting anonymously, confirmed that releasing CCTV footage of the two suspects had clearly been instrumental in finding the two suspects, either because someone had recognised them or, more likely he said, the suspects realised that they would be identified soon and chose to come out of hiding.
There are reports that the two suspects robbed a convenience store shortly after their images were released to the media.
“Whatever the reason, as soon as the FBI released the video footage, I knew that it was only a matter of time – perhaps a day or two at most – before they were identified,” he said.
The analyst, who is an expert in the use of the latest video forensic tools, was sceptical about claims made by producers of video analytic software that they could have prevented the Boston bombings by spotting the bombers before they struck.
The former president of Oncam Global, Dr James Ionson, claimed in a TV interview in America that the technology to identify suspicious behaviour already exists. He said he was frustrated that it had not been widely adopted by police departments. And Face First CEO Joe Rosenkrantz, interviewed on Bloomberg TV, claimed that video analytics could be used to identify thousands of people from CCTV images against databases of known individuals. Read related article here >>.
Much has been said in the past week about the use of images from mobile phones and camcorders. The analyst said that while they have a role to play, he was keen to put them into perspective. “On a mass event like that, the management of that much video material would be a nightmare,” he said. “You would use CCTV in the first instance to establish the movements of people and the timeline of events and then use crowd-sourced images – both video and still images – to fill in the gaps.”
Amateur video and still images could provide a different angle on a person of interest, but they would be difficult to identify in the mass of publicly-sourced material until you had gone through the process of establishing exactly where and when it was shot.
He was critical of the video that the FBI released of the two suspects, commenting: “The FBI did it horrendously.” He pointed out the ghost images that appeared in the video which are an example of frame blending, a video artefact which can result from converting time lapse video to play at higher speeds.[youtube]http://www.youtube.com/watch?v=M80DXI932OE[/youtube]
“It’s easily fixed,” he said. “I’m surprised that they rushed it out this way – I assumed they were delaying the release of the video in order to get it right.”
From his knowledge of how video is used in investigations, he said that the FBI would have much more video which they hadn’t released. The process of analysing this would be made easier by the suspects’ distinctive headgear. Most likely one team of analysts would look backward, working out where the suspects had come from while another would work forward to determine where they went after the bombings.
“Overall, this is a good case study for CCTV,” he concluded. “A camera sits there for years and years capturing almost nothing and then one day it helps us out.”