Biometric systems are developed for the detection of faces and gestures. The algorithms from real-time computer vision are optimised to a point where they are extremely robust. Therefore, they are even suitable for employment in difficult lighting situations. The modularly designed procedures can be used in systems for man-machine-communication, in access control and in teleconference systems.

The research area »cognitive systems« comprises developments for computer vision, man-machine-communication, 3D visualization as well as the development for smart camera systems for robot vision.

Face and Object Detection

Focus of our research and developments within the field of biometrics are detection technologies for faces and objects and also the fine analysis. Our library enables the detection of faces in video streams and still images. The succeeding fine analysis detects more details in the face and returns the gender and the detected facial expression. The expansion of the technology in order to recognize the age as well as other objects like hands, cars or logos are further steps towards visual perception similar to the human model.

Embedded Imaging

Complex processing – e.g. detection or adjustments – can be integrated into a camera by using embedded technologies for camera design. Therefore peripherical devices like computers for image processing will become obsolete. As a great advantage systems are becomming compact in size and flexible in use. The camera works like a stand-alone system. The concept of a »camera with embedded imaging« offers an ideal platform to integrate image processing functionality in a compact camera unit.

3D Reconstruction
3D modeling of faces and 3D reconstruction of objects based on camera images simplifies and accelerates cost and time consuming methods for computer graphics applications. No additional hardware like laser scanner or striped light or the tracking of markers is necessary to get a rough cut version of the object.