For automobiles ToF cameras are mainly used in three areas, one is driver fatigue monitoring in the cockpit, gesture recognition, and face recognition. The second area is solid-state lidar. ToF camera can be regarded as a kind of solid-state lidar, there are Flash type and single photon. The third area is the area of automatic parking.
Compared with the traditional single-camera fatigue driving monitoring, the ToF camera has many advantages. It not only has all the functions of a traditional camera, but also has head posture and line-of-sight analysis. The pure camera requires complex deep learning algorithms, a large amount of data collection, huge software development costs, and also consumes a large amount of GPU computing resources, which requires corresponding high hardware costs. Even so, the real-time performance is far inferior to ToF cameras.
Compared with structured light, robustness is greatly improved, especially at night, when facing car lights and street lights cause sharp changes in light, structured light is severely disturbed, and ToF is almost undisturbed. Both single camera and structured light require long-term calibration, especially structured light or stereo vision.However, ToF camera calibration is simple and fast, and does not require multiple calibrations, making it more convenient and intelligent to use.
