|Organisations are looking for ways to maximise resources and extract value from video security investment|
"There’s valuable information hiding within surveillance video. Extracting it can make a security department more efficient and provide a clear return on investment to their organisation." - Dror Irani, CEO and President, BriefCam - The Video Synopsis company.
Challenges of video surveillance
Video surveillance systems are a blessing to security. They are also a curse to those who have to justify their cost. Although video has become essential to safety and security, most organisations are still looking for ways to maximise resources, increase operational efficiencies and extract real value from their video security investment.
That’s unfortunate because surveillance video is rich with information that, if properly analysed, could be applied to any number of purposes. But video review and analysis has been a painstaking task with only an estimated 1% of video ever watched – and usually long after an incident has occurred. In most cases, the raw footage is disposed of after a certain period of time.
The challenge is to take that unstructured video and give it shape, much like an image search engine displays visual data. In the case of video, this can be done with video synopsis, a technology that cuts review time down to a fraction of the original run-time by detecting only moving objects that are then tagged according to clearly defined parameters, such as the time that an object enters or exits the frame, its colour, direction and path taken, size, speed, dwell-time, etc. Once indexed, they can then be called on, sorted, and presented according to parameters and hierarchies, e.g., largest vs smallest, fastest vs. largest, most red vs. least red, etc.
By presenting all moving objects on-screen simultaneously, while at the same time eliminating non-moving, extraneous information, (i.e. static background), days, weeks and even months of footage can be reviewed quickly, efficiently and accurately.
Video review and analysis
In an investigation scenario, the purpose is to get to the desired target in the shortest amount of time. Structured video needs to be processed only once; it can then be viewed in many different ways with the most relevant results presented first. For example, if investigators have some indications as to what they’re looking for (“he was loitering in this area”), objects extracted from the video and sorted by various characteristics would likely bring up the suspect first. In many ways, it is like a search engine where we have come to expect accurate answers on the first or, at the very most, the second page of results.
Analysis of Big Data over a period of time enables not only productivity but proactivity as well: with routine monitoring and alerts
This form of targeted video search and review saves on time and people-power. Rapid video review provides immediate insight into an organisation’s video data, investigation time is reduced dramatically, and higher quality evidence can be discovered.
With review time cut to the minimum, personnel can investigate more incidents both on-site and post-event, including those so-called “minor” incidents that generally go uninvestigated (e.g., shoplifting, false claims, etc.) because the cost of investigation outweighs the benefit.
Structuring video data also allows for dual-usage, with organisations deriving value from video for both security and non-security purposes. For example, businesses can utilise video metadata extracted for optimising operations, movement, space usage, market research, etc.
Moreover, analysis of Big Data over a period of time enables not only productivity but proactivity as well: with routine monitoring and alerts, users are able to discover events that previously would have gone undetected, giving them the ability to “know what you didn’t know before”. This is important specifically for large physical installations such as airports, mass transit, casinos, and the like, where continuous monitoring during off-hours and weekends is a must.
There’s valuable information hiding within raw surveillance video. Extracting, organising and leveraging it can enable better utilisation of the security investment, and highlight the role of the security department in supporting the organisation as a whole.