车内人员不安全行为的预警是减少交通事故的重要手段,然而现有的研究主要关注于驾驶员的危险行为检测,乘客的危险行为往往被忽视,因此文中提出了一种基于车云算力协同的驾乘人员危险行为分析算法。首先,通过基于粗粒度深度估计的检测和人员身份分析模型区分驾驶员与乘客;然后,通过基于ByteTrack 多目标跟踪的检测模型检测乘客是否越位,并通过头部姿态估计与疲劳状态检测的多任务联合模型检测驾驶员的危险驾驶行为;最后,将危险行为视频片段上传云端进行校验。实验结果表明,文中提出的基于粗粒度深度估计的检测模型和人员身份分析模型相比感兴趣区域法的F1-Score 提高了5. 1%,多任务联合模型中的头部姿态估计分支相比LwPoser 模型F1-Score 提高了5. 2%,疲劳检测分支相比YOLOv6s 模型平均准确率均值提高了2. 4%,且校验中使用云端高精度模型F1-Score 提高了7. 6%。以上结果证明了文中所提算法的有效性。

The warning of unsafe behavior inside the vehicle is an important means to reduce traffic accidents. However, existing research mainly focuses on detecting dangerous behaviors of drivers, while dangerous behaviors of passengers are often overlooked. Therefore, an algorithm for analyzing dangerous behaviors of both drivers and passengers based on vehicle-cloud computing collaboration is proposed in this paper. First, the driver and passengers are distinguished using a detection and personnel identity analysis model based on coarse-grained depth estimation. Then, a detection model based on ByteTrack multi-object tracking is used to detect whether passengers are out of their seats, and a multi-task joint model for head pose estimation and fatigue state detection is employed to detect dangerous driving behaviors of the driver. Finally, video clips of dangerous behaviors are uploaded to the cloud for verification. Experimental results show that the proposed detection model based on coarse-grained depth estimation and the personnel identity analysis model increase F1-Score by 5. 1% compared to the region of interest method, the head pose estimation branch of the multi-task joint model increases F1-Score by 5. 2% compared to the LwPoser model, the fatigue detection branch increases the mean average precision by 2. 4% compared to the YOLOv6s model, and the use of high-precision cloud-based models increases F1-Score by 7. 6% during verification. The above results demonstrate the effectiveness of the proposed algorithm.

TP391.4

国家自然科学基金资助项目(61976217)