Our facial recognition state-of-the-art solution is the most advanced solution for the most popular and performance challenging scenarios:
• Facial surveillance: Our real time facial recognition platforms build from the bottom-up and combined with the most advanced image processing algorithms to support the recognition engine makes ours the best performing solution worldwide.
• Facial investigation: Our facial investigator is the most advanced solution for law enforcement investigation, our pre-processing and post processing capabilities enable the recognition rate to be superior even with a low resolution and very hard scenarios images and videos.
Our solutions are currently being implemented in the Transportation, Medical, Banking, School throughout the HLS and security industry. Provides actionable intelligence to operators and security personnel Increases their effectiveness in detecting criminals and subjects on watch-lists.
Features that make the difference
We separate the reflectance and illumination of the original image, and then separate the illumination into geometric shading cues and lighting effects extrinsic to the face, such as cast shadows. We then recombine the reflectance with the geometric shading information to produce a natural-looking image similar to what would be observed in favorable, controlled illumination conditions and is thus better-suited for recognition.
The limitation of Face Recognition systems is a requirement that people look straight into the camera to be identified. In face surveillance and especially in its covert deployments, this requirement is untenable.
IDT Face Middleware has developed technology that exploits the uniqueness of human face and leverages the three-dimensional nature of the world to mitigate poses and facial expressions thus eliminating the need to interrupt individuals’ natural movement with requirements to pose for picture or moderate a smile.
Different lightning such as neon or led direct light can cause hot spots across the face that will have significant effect on the recognition accuracy.
The biggest challenge is to remove those hot spots without demonstrating any changes that will affect the recognition, meaning removing the hot spots without changing the biometric features across the face.