Look up. They’re all around us.
Wherever you cast your eyes, there’s a security camera, recording your every movement, day in and day out. But don’t worry. Less than 2% of that video will ever be watched and then only if something bad has occurred. And usually long after the fact.
Which poses the question: if almost no one’s making use of recorded video, what’s driving the exponential growth in surveillance and security camera installation around the world?
There are several reasons why surveillance has been growing: better security camera quality coupled with lower camera costs, security consultant recommendations, insurance premium discounts, and a general sentiment that it is better to have videotaped and not watched than never to have videotaped at all.
Surveys have also shown that, in general, people feel that video surveillance – at least in public spaces – is less invasive than surveillance of phone calls or emails. Surveys have also shown that public cameras give people a feeling of safety.
But again, if no one is making good use of the video, are those feelings justified?
Generated from pure research
BriefCam came into being in response to the problem of “too much video, not enough eyes to watch it.” Like many an Israeli start-up, BriefCam grew from pure research. Prof. Shmuel Peleg of the Hebrew University, a world-renowned expert in computer vision, was trying to solve a particular problem. It required taking the video frame, separating static backgrounds from dynamic moving objects, then databasing and indexing those objects. The resulting technology – Video Synopsis – enabled users to call on those objects and display them simultaneously on the background – even though they occurred at different times.
The summary video was significantly shorter than the original, with hours of video reduced to minutes of video review runtime.
In December 2007, Video Synopsis was licensed for commercialization through Yissum, the technology transfer arm of the Hebrew University to BriefCam. This arm of the University is attempting to revolutionize the way surveillance video is reviewed.
The video reality
I don’t use the term “revolution” lightly. To date, video review has been so painstaking and time-consuming that the industry has come to accept that there will never be “enough eyes to watch it all” with only the most critical of crimes ever being investigated. Yet security cameras continue to proliferate, further exacerbating the issue.
There are solutions and tools. On the professional level, Video Management Systems (VMS) manage thousands of security cameras and employ identification and analytical tools, many of which attempt to automate the video review process with varying levels of success; analytics often fail at differentiating between events that pose a threat and those that do not. For example, a human being can tell in an instant whether a slap on the back is friendly or aggressive. A software algorithm will most likely not be able to sense this ambiguity.
BriefCam maintains that video review cannot be fully automated and that human operators – their eyes and minds – must always be engaged.
Fast video review – either for post-event investigation or for real-time investigation – makes it possible for people to go over video rapidly and take action as needed. Investigations that once took protracted periods of time can now be reviewed in a matter of days and even hours. It took months to review the video used in the 2005 London bombing investigation and weeks to review 5000 hours of video from the Vancouver riots. All of that could change soon.
Video value extracted
Day to day, the full value of the investment in a surveillance system could be maximized if a higher percentage of recorded video were reviewed. There would be a greater likelihood of discovering events previously unnoticed. These include shoplifting, graffiti, staged “slip and fall” accidents (used to dupe insurers), and a range of other petty crimes currently deemed too insignificant to warrant investigation.
Take the example of shoplifting. A storeowner knows exactly when a robbery has occurred – and doesn’t need BriefCam to tell her when the till was emptied. But the storeowner could have earned 10 times the amount of money stolen from the till if she had learned to review footage effectively. A periodic review of the day’s video could provide valuable information that would, within a fairly short time, bring a return on investment.
Commercial enterprises can also leverage their video’s value by applying it to business intelligence. For instance, they can examine customer behavior patterns or monitor employees to boost profitability.
The issue of ROI will become even more critical as private users acquire surveillance systems. All research indicates that the consumer/SMB surveillance market is growing rapidly with exponential growth predicted in the coming years. These are people with family and property to protect, and no time to spare going over hours of video. They’ll want Video Synopsis accessible over computers, mobile phones, and tablets. They’ll want it on-demand or, in the case of an alarm, they’ll want to receive an immediate 5-second Synopsis of the 5 minutes recorded prior to the alert. They will want their investment in video to be meaningful. For them, rapid video review will be essential.
Get to the point. Fast
Our society lives with security cameras, large and small. We accept their presence in the public arena out of a feeling that they safeguard us and that, should something go wrong, the recorded video will provide evidence. This perception is flawed as most video goes unwatched or is watched only long after the fact. For this reason, rapid review is essential to our videoed society. It meets the public’s expectation that video will be reviewed when necessary and evidence pinpointed as fast as possible, so that action can be taken or preventative measures made on their behalf. It also unleashes video’s full potential, whether for saving time, for saving money, and, most importantly, for saving lives.
Links in this article:
Poll Finds Strong Acceptance for Public Surveillance
http://www.nytimes.com/2013/05/01/us/poll-finds-strong-acceptance-for-public-surveillance.html
London bombers staged ‘dummy run’
http://news.bbc.co.uk/2/hi/uk_news/4263176.stm
How the FBI Will Analyze Thousands of Hours of Boston Bombing Video
http://www.popsci.com/technology/article/2013-04/how-analyze-thousands-hours-boston-bombing-video
