The system adopts monocular computer vision algorithm based on deep learning neural network and stable and fast vehicle-level FPGA platform, which can realize the functions of front collision warning (FCW), pedestrian collision warning (PCW), and lane departure warning (LDW). At the same time, the system can work in all kinds of weather and road environments, and can perform full-angle full attitude detection on various types of targets.

The system realizes driver fatigue warning and identity identification by detecting driver's head posture, facial features, and opening and closing eye state to ensure driving safety; Based on deep learning computing vision technology, the system can operate in a variety of complex environments (such as wearing glasses / light-colored sunglasses, complex lighting, passengers at night or in the rear)