摘要
在先进驾驶辅助系统中,基于视觉的行人检测只能对摄像头视野范围内的无遮挡行人进行检测,并且易受天气的影响,在极端天气下无法工作。针对视觉检测的缺陷,提出了一种利用超宽带(Ultra Wideband,UWB)通信模块检测行人位置信息的方法,并对其进行卡尔曼滤波以减小误差,同时将得到的行人位置信息与基于视觉的行人检测信息融合,设定匹配规则防止多报警和误报警的情况。测试结果表明,融合方法能够增加对视野外以及被遮挡行人的危险预警,提高了行人检测预警系统的可靠性。
For advanced driver assistance systems(ADAS),the traditional vision-based pedestrian detection techniques can only detect unobstructed pedestrians within the camera’s field of view,are susceptible to the weather and cannot be applied under extreme weather conditions.Therefore,a method to acquire the pedestrian-position information by using Ultra Wideband(UWB)module was proposed.The Kalman filtering was applied to reduce the error and then the obtained pedestrian-position information and the vision-based pedestrian detection information were fused to prevent false alarms or receiving multiple alarms for every event.The test results show that the proposed fusion method can increase the accuracy of risk warnings for the pedestrians obstructed or outside of the field of view,thus improving the reliability of the pedestrian detection system.
作者
曹立波
向国梁
张乐祺
廖家才
王秋丽
CAO Libo;XIANG Guoliang;ZHANG Leqi;LIAO Jiacai;WANG Qiuli(Key Laboratory of Advanced Design and Manufacturing for Vehicle Body,Hunan University,Changsha 410082,China)
出处
《汽车工程学报》
2020年第4期235-242,共8页
Chinese Journal of Automotive Engineering
基金
汽车车身先进设计制造国家重点实验室开放基金“汽车智能辅助驾驶多合一集成技术”(31715013)。
关键词
先进驾驶辅助系统
行人检测
融合预警
超宽带
advanced driver assistance system
pedestrian positioning
fusion warning
ultra wideband
