Ipsotek Releases AI-Based Module for Social Distancing Awareness
Ipsotek has released an analytics-based proximity detection module for social distancing awareness that is designed to alert security teams or management to personal proximity that may pose a health risk in public or work spaces.
Ipsotek’s VISuite AI tracks the GPS co-ordinates of all persons in the camera FOV in real time. VISuite AI is able to precisely calculate the distance between each person and generate snap alarms immediately the distance falls below 2m. It also accurately measures the time people were less than 2m apart, so if required, alarms are only generated for breaches over a prescribed period, E.g. 3 seconds, to identify people consciously engaging.
Ipsotek’s patented Geospatial algorithm automatically follows individuals across multiple cameras. This enables rapid contact tracing during forensic reviews, when it is necessary to identify those who came into contact with a person found to have contracted the virus. The top image shows the proximity of members of the public to one another. With the red circles pinpointing individuals who are within 2m of another person, thus indicating a breach of the social distancing guidelines.
Dashboards may be customized for visualizing the metadata gathered in real-time or for reviewing historical data. The reporting tools included in VISuite AI are extremely flexible and intuitive and allow operators to view multiple reports simultaneously. The example above is displaying a social distancing dashboard with 4 sections; violations over time, the total number of violations per day, a heat map representing the location of the violations and the average, minimum and maximum distances maintained between individuals.
“One of our customers approached us and asked if it were possible to highlight distancing between individuals in order to promote social distancing awareness,” Ipsotek’s CTO, Dr. Boghos Boghossian said. “Using our artificial intelligence video analytics (AIVA) platform VISuite AI, we trialled different attribute detection techniques and identified a reliable method for determining the distance between moving objects.
“I have personal experience of how this pandemic has affected people, so when when this challenge was presented, we did not hesitate in mobilizing our R&D team to determine an accurate method for defining the relative distances between moving objects. The proximity detection module will assist in determining areas where social distancing is not being observed, while also benefiting essential businesses who need to ensure their staff are protecting themselves and each other during their working day.”
Contact: [email protected] or +65 9163 0888