Biometric surveillance of vehicle occupants in unconstrained environments is rife with a host of challenges. In particular, the complications arising from imaging through vehicle windshields provide a significant hurdle. Distance to target, glare, poor lighting, head pose of occupants and speed of vehicle need to be addressed to ensure high accuracy recognition rates. To mitigate the issues present at image acquisition a custom, multi-unit camera system was constructed to provide the best possible face templates to enhance the performance of recognition software. System ruggedness and data throughput were of additional importance given the design use case of identity confirmation at a security portal. Face recognition software implementation includes high accuracy RCNN face detection and the use of a novel HDR GAN to provide high quality faces for evaluation.