摘要
自智能交通系统出现以来,汽车驾乘员的安全带检测一直是备受关注的研究课题.依据城市道路的交通卡口监控数据,研究一种基于深度学习的汽车驾乘人员安全带检测算法,能够准确识别驾驶员是否佩戴安全带.通过对卡口图片进行人工标定,并运用深度学习方法训练两个检测器和一个分类器,最终实现安全带的快速定位和分类.本文提出的方法在城市道路卡口采集的图像上检测效果较好.
With the development of intelligent traffic systems （ITS）, the detection of drivers＇ safety belts attracted a lot of attention. In this paper, we studied a safety belt detection algorithm based on deep learning, which could help accurately identify whether or not the driver wore the safety belt. The method could locate and classify the safetey belt by using deep learning to train the two detectors and a classifier. The method has a better detection effect on traffic monitoring images collected in urban roads.
出处
《中国计量大学学报》
2017年第3期326-333,共8页
Journal of China University of Metrology
基金
浙江省自然科学基金资助项目(No.LY15F020021)
浙江省科技厅公益性项目(No.2016C31079)
关键词
安全带检测
目标检测
深度学习
图像分类
智能交通
safety belt detection
object detection
deep learning
image classification
ITS
