Facial Recognition Deployment Framework for testing in real world settings developed by SecurLinx
25 September, 2007
category: Biometrics, Transit
Due to mixed performance results from designing and deploying facial recognition systems in difficult environmental conditions, SecurLinx has developed a Framework for effective testing in the laboratory and in real world settings. Settings such as sport venues or transportation hubs which have variable lighting can be tested. The field toolkit can measure these changes in detail and provide the necessary data to improve performance at the site with supplemental lighting, algorithm selection or both. The laboratory version of the Framework supplies detailed test data to developers to aid in algorithm performance improvements.
SecurLinx Introduces Facial Recognition Deployment Framework
Morgantown, WV — SecurLinx Corporation introduced its Facial Recognition Deployment Framework at the 2007 Biometrics Consortium Conference.
SecurLinx‘s focus is on designing and deploying facial recognition systems in difficult environmental conditions. As a result of the mixed performance results from these endeavors it became evident that a much more rigorous analysis of algorithm performance in relation to environmental conditions would be necessary to achieve effective deployments in real world settings. This internal need to test our own designs became the driver for the development of the Framework. It became evident that this solution could provide significant benefit to others in the industry facing the same issues.
“We are able to test our product performance in the same environmental conditions in which it will be deployed,” stated Barry Hodge, CEO of SecurLinx. “This will allow us to set our customer’s expectations appropriately while delivering the highest level of performance available in the marketplace,”
The patent pending technology consists of three major components that allow developers, users and integrators to test the performance of complete facial recognition systems in the laboratory and in actual field settings. The system enables repeatable, objective testing to create a performance profile and measure environmental characteristics.
“The physical settings where many users wish to deploy facial recognition systems are often an environmental challenge do to the chaotic nature of the location. Areas such as entrances to sports venues or transportation hubs have extreme variability in lighting conditions throughout a day and night,” explains Steven Rehfeldt, Chief Technology Officer. “The field toolkit can measure these changes in detail and provide the necessary data to improve performance at the site with supplemental lighting, algorithm selection or both. The laboratory version of the Framework supplies detailed test data to developers to aid in algorithm performance improvements.”