Prevents duplicate counting of the same person using face recognition technology. Employees can be filtered out from traffic counts.
SAFR from RealNetworks, Inc. a leader in high accuracy, low bias facial recognition, today announced a collaboration with Dains, a Korean company that specializes in unmanned people and asset counting solutions.
As part of the collaboration, SAFR is providing its AI-based computer vision technology to increase the accuracy of people counting to prevent duplicates of customers and employees.
The combined solution was recently deployed at the National Memorial Hall of the Korean War Abductees in Paju
, Gyeonggi-do. Since there is no entrance fee for the memorial, accurate statistics on the number of unique visitors were required. Dains developed an unmanned counting system with a feature to prevent counting duplication by utilizing SAFR's face recognition technology
With ceiling mounted camera counting systems, it is difficult to avoid duplicate counts when the same person enters or exits the premises multiple times. Employees entering and exiting would also skew count accuracy. Using SAFR facial recognition, the system can ignore multiple re-entries from visitors while opted-in staff can be removed from the total count entirely. Ceiling mounted cameras can rarely be used to identify individuals due to their limited field of view. SAFR facial recognition enables cameras to be installed at standard surveillance mounting heights so as to reliably capture individuals and events.
Dains plans to expand the unmanned counting product using SAFR's face recognition technology to additional markets. It also plans to offer options for analyzing visitor gender and age, enhanced employee attendance management, and access control.