驾驶员不良情绪状态监测及舒缓系统开发

本系统集监测与舒缓功能于一身实现驾驶员不良情绪状态实时、非接触式地监测及舒缓.该系统首先使用肤色模型算法迅速定位人脸,然后利用累积差分帧和 Hough 变换等实时图像处理技术进行检测、跟踪眼睛和嘴巴的状态并提取相对特征参数,进而与表情模板匹配,以相似度衡量相似程度并最终判定驾驶员的实时情绪状态.如果判定的结果为不良情绪状态,系统将自动播放与之相对应的舒缓曲目,最终达到舒缓驾驶员情绪、保证安全行车的目的.

Research on the Monitoring and Mediating System of Driver＇s Negative Mood State

The system integrated monitoring and mediation into one body that can effectively monitor and mediate driver＇s negative mood state, in time and without contact. At first, the driver＇s face is localized using the skin color model, then, the technology of the real-time images is used to detect the transform rack and analyze the state of his eyes and mouth and pick-up the characteristic parameters of the eyes and mouth by transform. Then match the expression pattern, determine the real-time driver＇s mood state according to the similarity degree. If the determination result is a negative mood state, the system automatically plays corresponding soothing track. Finally, the ultimate goals of easing driver＇s negative mood and ensuring safe driving is achieved.