Face Detection and Recognition based on Fusion of Omnidirectional and PTZ Vision Sensors and Heteregenous Database

Large field of view with high resolution has always been soughtafter for Mobile Robotic Authentication. So the Vision System proposed here is composed of a catadioptric sensor for full range monitoring and a Pan Tilt Zoom (PTZ) camera together forming an innovative sensor, able to detect and track any moving objects at a higher zoom level. In our application, the catadioptric sensor is calibrated and used to detect and track Regions Of Iinterest (ROIs) within its 360 degree Field Of View (FOV), especially face regions. Using a joint calibration strategy, the PTZ camera parameters are automatically adjusted by the system in order to detect and track the face ROI within a higher resolution and project the same in facespace for recognition via Eigenface algorithm. Face recognition is one important task in Nomad Biometric Authentication (NOBA 1 ) project. However, as many other face databases, it will easily produce the Small Sample Size (SSS) problem in some applications with NOBA data. Thus this journal uses the compressed sensing (CS) algorithm to solve the SSS problem in NOBA face database. Some experiments can prove the feasibility and validity of this solution. The whole development has been partially validated by application to the Face recognition using our own database NOBA.