基于动态时间规整的酒驾识别研究

On Drunk Driving Detection Based on Dynamic Time Warping

针对酒驾行为识别问题,提出了一种基于动态时间规整的识别方法,利用智能手机采集到的侧向和纵向加速信息,动态匹配酒驾的行为模式,智能判断是否处于酒驾状态.实验结果表明,该方法切实有效,且有一定的应用价值.

In view of drunk driving detection, a new algorithm based on dynamic time warping is proposed, combining the data collected by the longitudinal and lateral acceleration of smartphone, dynamically matching the drunk - driving schema and finally judging whether the car is under the state of drunk driving. The experimental results show that this method is effective and practical.