基于人脸3D模型的驾驶人头部姿态检测  

Detection of Driver's Head Posture Based on 3D Face Model

作　　者：张波[1] 王文军[1] 成波[1] 

机构地区：[1]清华大学,汽车安全与节能国家重点实验室,北京100084

出　　处：《汽车工程》2016年第1期53-60,71,共9页Automotive Engineering

基　　金：国家863计划项目(2012AA111901)资助

摘　　要：为对驾驶过程中的驾驶人注意力分散行为实施监控和预警,提出了一种基于人脸3D模型的驾驶人头部姿态检测方法。该方法使用ASM算法与ASEF算法定位驾驶人的正面人脸;在此基础上,基于通用人脸3D模型与正面人脸计算当前驾驶人的人脸3D模型;以正面人脸3D模型为姿态度量的基准,运用Pn P算法实时解算驾驶人的头部姿态,并进行姿态有效性判别;同时提出了一种特征点更新的方法,以提高姿态解算的鲁棒性。实验结果表明,该方法能够在驾驶室环境中有效检测出驾驶人的头部姿态。For monitoring and forewarning the distraction behavior of driver during driving, a detection method of driver＇s head posture based on 3D face model is proposed. The method utilizes active shape model algo- rithm and average synthetic exact filter algorithm to locate the front-view face of driver. On this basis, the 3 D face model of the specific driver is calculated based on the general 3D face model and front-view face. With the frontal 3D face model as the base of posture measure, PnP algorithm is adopted to calculate the driver＇s head posture in real time, with the posture effectiveness discriminated. Meanwhile a scheme for updating characteristic points is also put forward to improve the robustness of posture calculation. The results of experiments show that the method proposed can effectively detect the head posture of driver in cabin environment.

关 键 词：驾驶人 头部姿态 机器视觉 人脸3D模型 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] U463.6[自动化与计算机技术—计算机科学与技术]