How viisights detects suspicious activity

Virtually every business, organization and municipality around the world employs some form of video technology to enhance security, and increasingly for business intelligence to enhance operations and services. Industry analyst HIS Markit estimated that there may be as many as 1 billion CCTV cameras in use by the end of 2021. Yet while the number of cameras deployed seems to keep growing every year at a staggering pace with improved image resolution and overall capabilities, their primary purpose is for live monitoring and to record visual data for forensics. Which means that unless these cameras are being diligently monitored by security or law enforcement professionals, they have little effect in preventing potentially threatening events from unfolding.

While some improvements in sensor technology allow video systems to detect and alert system administrators when some levels of suspicious activity take place, such as when an object is left behind or object movement and direction, they still require live monitoring to analyze the situation and initiate action. Being able to analyze the behavior of people and vehicles based on their specific actions, surrounding area and the events in which they are participating is the pathway to a whole new level of video understanding and behavioral recognition video analytics

viisights sets the benchmark with innovative behavioral recognition video analytics that can detect and analyze suspicious activities.

We’ve harnessed the power of Artificial Intelligence (AI), machine learning, and deep learning, to detect and understand a wide range of human and vehicle behaviors to automatically identify abnormalities and potential threats. Already deployed at numerous locations around the world today, viisights behavioral recognition video analytics are helping security operations and law enforcement agencies detect suspicious human activity using our video recognition system before potentially dangerous events take place or would otherwise go undetected.

Event prediction is one of the main goals of viisights’ unique time series analysis. Prediction can play an effective role for appropriate decision making in different applied areas. WISE near real-time, temporal analysis technique, which is based on behavioral recognition, enables suspicious behavior detection which could potentially lead to other events of interest.

The simplest use case is loitering, which in some contexts, like near an ATM, can be indicative of planned fraud or theft. Driven by behavioral understanding algorithms, WISE can predict even more complex situations. It can detect an event of continuous unrest in a crowd. Such behavior is, in many cases, a predictive indicator for an upcoming violent activity. Through early detection, the user can take actions to prevent any potential violence. Though these use cases are timely and relevant, the potential for the predictive capabilities of WISE is far reaching and promising.

Suspicious Behavior Detection Showcase Video

Here are some of the suspicious activities that viisights innovative behavioral recognition video analytics uniquely identifies and analyzes:


Contextual Loitering

Loitering while waiting for a bus or friend to arrive may be completely innocent, but individuals spending excessive amounts of time near ATMs, or in parks, building lobbies and parking facilities may indicate a potential problem. viisights intelligent behavioral recognition video analytics automatically detect such activities and issue alerts in real time, dramatically increasing contextual awareness for personnel to investigate the situation before an incident occurs

Object Abandonment

Abandoned bags or parcels are typically not of much concern – perhaps the result of a distracted traveller – or perhaps intentionally placed with ill intent. viisights intelligent behavioral recognition video analytics can quickly identify an abandoned object and alert security or law enforcement personnel of its presence so they can investigate the situation and determine who may have left it there.

Entering or exiting a vehicle

People entering and exiting a vehicle at a taxi line or a hotel is typical behavior, but when such activity takes place in front of a bank, jewelry store or school in the middle of the night, there is cause for concern. viisights behavioral recognition video analytics can automatically alert authorities as suspicious activities are taking place so that they can address the situation before an event escalates.

Appearance and Disappearance of Objects

Some video cameras can detect movement, and even when an object is left behind, but few, if any, can detect and let you know when an object has been removed from a scene. viisights behavioral recognition video analytics can identify objects placed in specified areas, and when objects have been removed – a security advantage that even the most highly trained security personnel can struggle with when monitoring live cameras for extended periods of time. viisights behavioral recognition video analytics never blink or doze off, automating the detection process with extreme precision.

Person falling/on the Ground

There are many reasons why people may be on the floor, but they may just be tying their shoes, or picking up an item they dropped. They may also be in distress from a medical condition, or pushed to the floor, or worse. In any case, viisights behavioral recognition video analytics provide the contextual analysis to identify exactly what is happening, alerting the proper personnel to investigate if they need to initiate remedial action.

Hidden Face (person with/without a mask)

One of the first things a potential criminal will do before committing a crime is to cover their face. And with the onset of the COVID pandemic, people covering their faces is also required in many different venues around the world. This presents a new challenge for security and law enforcement personnel. viisights unique video understanding technology can determine when an individual’s face is covered, when it’s not covered and should be, and when a suspicious person may be intentionally hiding their face or walking backwards toward a video camera. viisights has face detection and analysis all covered.

Person Holding Weapon

Weapon detection by way of behavioral analysis is a powerful way to not only detect threatening and potentially dangerous situations, but also eliminate false alarms. By using body movements and behavior to determine the existence of a weapon, rather than pixels or shapes, viisights is able to alert operators when a weapon is being brandished. This ensures that holstered weapons, such as those on the hip of a law enforcement officer, do not trigger an alert.

Person Movement Type

Running in areas where walking is normal, such as a department store aisle or shopping mall can be cause for concern. Climbing or jumping can be normal in a park or during a game, but over a fence they can indicate a troubling activity. viisights behavioral analytics apply the intelligence and power of AI-driven contextual understanding to detect and identify such incidents as they unfold so they can be addressed accordingly – or not at all.