Design of 2D gimbal face localization and fatigue driving detection system

Aiming at the traditional fatigue driving detection system with a single function, unable to locate the face, greatly affected by external lighting factors and high hardware requirements, this design proposes a system that uses OpenMV equipped with a CMOS camera as the main controller, recognizes the face contour by using Haar classifiers, and controls the two-dimensional gimbal to realize the face localization function, with the MCU as the slave, and utilizes the MCU to read the blinking frequency per time unit to determine the fatigue driving. The microcontroller is the slave, using the microcontroller to read the blinking frequency captured by OpenMV per unit time to judge the fatigue driving, when the blinking frequency is higher than 30 times/min, the microcontroller triggers the sound and light alarm. At the same time, the neck and waist signal analysis module is added to collect more comprehensive driver fatigue information. For complex road conditions such as sharp curves, up and down slopes, the microcontroller can pre-judge in advance through the positioning module to remind the driver in the form of alarm. The experiment proves that the system adopts lightweight design scheme, applicable to the field of in-vehicle electronics, both with the advantages of diversified functions, high measurement accuracy, intuitive display, stable operation, etc. The combination of 2D PTZ face localization and fatigue driving detection in the mode of master-slave improves the real-time and robustness of the control system, which is useful for the improvement of fatigue driving detection by using machine vision.