AI AND VIDEO ANALYTICS BLOG
Video Surveillance & Physical Security Industry Viewpoints
September 5th, 2019
Author: Enav Perez

How Financial Institutions Can Maximize Video Intelligence for Security and Beyond

Physical Property Security Matters

Large financial firms—such as mortgage companies, international and national banks, credit card companies, mutual fund companies and the like—manage vast quantities of personal and financial data. For smaller banking organizations, sensitive data and critical infrastructures might be maintained on-premise, but for larger institutions with multiple branches, these are stored in offsite centralized data centers. Data centers are particularly essential for financial institutions: A systemic failure of a financial organization’s data center could be catastrophic because it could compromise personal information and make it difficult or impossible for their customers (consumers or businesses) to conduct financial transactions.

Over the past several years there has been much attention on digital data security, which is perfectly understandable given the high privacy stakes and constantly evolving cyber threat landscape, with hackers seemingly always one step ahead of security researchers. However, it’s equally important for financial organizations to protect their physical sites such as business offices, branches and data centers.  For financial institutions, the security emphasis tends to be on maintaining situational awareness, detecting threats to the facility such as unauthorized intrusion, property damage, and even theft. Security staff typically monitor building entrances/exits, tracks pedestrians and vehicles, and grants or blocks access to certain parts of a building, with video surveillance acting as a key security enabler.

Harnessing the Power of Video Intelligence

Although video surveillance is often adopted as an adjunct to security staff, video surveillance systems have limited impact and most video data is under-utilized. This is often because of the following factors:

  • Video cameras produce more footage than security staff have time to review in reality;
  • Actively monitoring an entire network of cameras (especially a large network, like those of major financial companies) in real-time is unrealistic;
  • Human operators are prone to distraction and error.

Video content analytics software is one way finance and banking organizations can overcome these challenges and extend the functionality of their existing video surveillance infrastructure. Powered by Deep Learning and Artificial Intelligence, video analysis software processes video, identifies objects in the video footage (people, vehicles, and other objects), and indexes them so that footage can be easily and quickly searched and analyzed. From a security standpoint, this technology is valuable because it allows security teams to quickly search, review and take action based on video. They can pinpoint relevant details, extract crucial data and accelerate investigation of incidents with the information mined from the video footage.

Video investigation technology also allows security teams to set up real-time alerts to enhance situational awareness and trigger action by security staff. When the video analysis software detects any loitering, activity by entryways or any other potentially suspicious behaviors, it can configure alerts to be triggered to prompt security staff to assess the activity and determine the appropriate response in real-time, as the situation develops. For example, the security team could configure count-based alerts; if there are five people detected in the building in the middle of the night when only two are authorized for that hour, security staff can be informed of the additional entries detected and investigate. Similarly, if movement is detected at off-hours in a restricted area of a building or parking lot, the VCA system can alert security staff so they can respond.

Real-time alerts can also be set up based on facial recognition (where allowed); for example, a watchlist of people who are authorized to be in the building at 2 am can be compiled, so that alerts can be triggered when people that do not match faces on the list are detected. It can also notify staff when a known suspect enters the premises. For post-event investigation after an incident occurs, video analysis software enables the use of both facial recognition – to search for a particular face – as well as object extraction and identification – to detect the person of interest’s other identifying characteristics, such as gender or color of clothing.

Observe Trends to Improve Security and Operational Efficiency

With VCA technology, video data can also be aggregated over time and then visualized as reports, dashboards and heat maps, to reveal trends that operations and security staff can use to understand pedestrian and vehicle traffic patterns across its data centers, business branches or office facilities – locally, nationally and internationally. Video analysis reveals where people tend to congregate and average dwell times for different areas. By referring to those trend reports, corporate management can improve efficiency, performance, and security.

Financial institutions already spend significant resources on site security. But they can further leverage their existing investment in video surveillance by deploying video content analytics software, which transforms video content into actionable intelligence, reducing the overall costs of security while streamlining its enablement. By maximizing investments in legacy video surveillance technology with video content analytics, financial institutions can investigate crimes and suspicious behavior, respond to developing situations, increase situational awareness and security, while also enhancing operational efficiency for all its business groups and driving scalability as the organization expands.