WASHINGTON, December 7, 2016 — Any time you walk into a bank, peruse a department store or stroll through a parking structure, chances are you are under the watchful eye of at least a few surveillance cameras.
For years, video surveillance has been a tool used by businesses and other organizations, mainly for security purposes. Although it has been useful for monitoring and summoning evidence in the event of a crime, however, video surveillance does have its drawbacks.
Sorting through archives of 24/7 footage can take hours. Worse, the attention span of those monitoring screens and reviewing footage often falls far short of the level of what’s needed to maximize the value of the tool. Indeed, video has become such a powerful and pervasive medium in the past decade that mere security surveillance does not seem to represent the full extent of what this tool is capable of offering to businesses.
In recent years, we have seen impressive advancements in video surveillance technologies, resulting in a new area of study known as intelligent video analytics (IVA). Following the industry trend toward the use of big data, video analytics software systems use surveillance cameras to gather and analyze visual data from within the walls of brick and mortar stores for both business and security purposes. The resulting vast array of possibilities provides evidence that we have been failing to exploit the possibilities of video surveillance for years.
A recent Rutgers University study outlined some of these possibilities.
Advantages for Business
IVA software can be programmed to count the number of patrons entering and leaving a store throughout the day. The data can be used to help business owners correlate store traffic with financial information, determine how long patrons typically spend in the store and identify high-traffic times to optimize hours of operation and staffing levels.
Video analytics can track locations where people spend their time while in a store. Understanding things like where customers spend the most time, where they go first and how they navigate through the rest of the store can lead to better and more efficient store layout designs. The data gathered can also help business and marketing staff to make more informed decisions on key matters such as the placement of displays and advertising within the store.
Similar to movement tracking, visual interface software identifies landmarks within a store such as a checkout stand or a prominent new display, enabling the analytics software to analyze and report on how customers interact with these landmarks or displays. This feature is useful for pinpointing individual successes and and problems within individual stores.
Visual analytics can provide key demographic information. For example, IVA software can now identify customers’ age, gender and ethnicity—key information for any retailer planning its upcoming seasonal offerings.
Advantages for Security
IVA programs can recognize suspicious activity such as an individual trying to conceal medium or large objects under his or her clothing in an attempt to shoplift. This is accomplished by means of an algorithm that compares the silhouette of the individual to the normal shape of the human body. When the software notes something highly suspicious in this regard, store personnel are alerted to the possibility of theft.
Other Anti-Shoplifting Measures:
In addition to recognizing people who appear to be shoplifting, some video analytics software also recognizes the facial features of questionable individuals. This feature can be extremely useful, particularly if an individual previously identified as a shoplifter is now re-entering a company store. In addition, a target retrieval feature allws IVA systems to track a suspect’s location within the store, making it easier for security to monitor and address the situation.
“Sweethearting” occurs when an employee secretly gives unauthorized discounts or free merchandise to friends or family members—a widespread problem in the retail world. IVA systems have been developed to detect at least some of these inside operations by setting event rules for what a transaction should look like. For example, during a standard, in-store transaction, a cashier should be located behind the counter while a customer is generally in front of the station. Additionally, each item in the transaction should be scanned. If the software notes the absence of key elements such as these, the system notes the transaction is suspicious, indicating that sweethearting or other illegal transactions may be in play.