Realization and Optimization of Embedded Driver Status Detection System

This paper presents a novel driver status detection algorithm, which can be processed in embedded platform in real-time. The proposed method is based on Adaboost and dynamic modeling algorithm. Compared to the traditional active infrared radiation method, our system employs a safer passive way and the algorithm is more robust to the various illuminations. There are two main contributions: 1) it mixes up the single eye and eye pairs detectors together and presents an adaptive detection region Adaboost eye detection algorithm, improving the detection rate and speeding up the eye detection process; 2) it presents a dynamic eye modeling tracking algorithm which is based on the Gaussian mixture model. The tracking algorithm can automatically extract image intensity distribution of the driver s eye region, thus can model and track eyes of different drivers. Experiments on several public databases and driving sequences taken in real car show that the proposed method can detect the driver status precisely and satisfy the real time processing requirements.