摘要
环境视觉感知技术是智能车获取外界环境信息的主要手段,是智能车进行自主驾驶的前提条件。在综合分析以不同检测对象为目标的检测技术方法的基础上,给出了环境视觉感知研究的主要内容,讨论了研究涉及的关键技术:车道检测技术、车辆检测技术、行人检测技术、交通标志检测技术,并对每种检测技术各子技术的优点和不足之处进行了分析。在总结全文的基础上,提出未来智能驾驶中环境视觉感知技术的主要发展趋势包括:基于多特征融合的环境视觉感知技术的发展;复杂环境下多种环境感知技术的融合应用;三维图像在图像处理方面的应用;改进的卷积神经网络(CNN)方法在图像处理方面的应用。
Environmental visual perception technology is one of the main means for intelligent vehicles to obtain external environmental information and the precondition for autonomous driving of intelligent vehicles.Based on the comprehensive analysis of the detection technology and methods targeted at different detection objects,the main contents of environmental visual perception research are provided,and the key technologies that the study involves including lane detection technology,vehicle detection technology,the pedestrian detection technology,and traffic signs detection technology are discussed.The advantages and disadvantages of each sub-technique of each technology are analyzed.Finally,on the basis of summarization of the paper,the prospect of unmanned environment of visual perception technology in the future development direction is given:the development of the multi-feature fusion based environment visual perception technology;the fusion application of several environment perception technologies under complicated conditions;the application of the three-dimensional image in picture processing;the application of the improved CNN method in picture processing.
作者
陈政宏
李爱娟
邱绪云
袁文长
葛庆英
CHEN Zhenghong;LI Aijuan;QIU Xuyun;YUAN Wenchang;GE Qingying(School of Automotive Engineering,Shan Dong Jiaotong University,Jinan,Shandong 250357,China)
出处
《河北科技大学学报》
CAS
2019年第1期15-23,共9页
Journal of Hebei University of Science and Technology
基金
国家自然科学基金(51505258
51775268)
山东省自然科学基金(ZR2015EL019)
山东省重点研发计划项目(2016GNC112010
2017NC212002)
山东省农机装备研发创新计划项目(2017YF013
2018YF017)
关键词
车辆工程
视觉感知
车道检测
行人检测
车辆检测
交通标志检测
vehicle engineering
visual perception
lane detection
pedestrian detection
vehicle detection
traffic signs detection
