摘要
为了实现自动驾驶汽车障碍物识别判据的量化,提出一种基于椭圆包络曲线集自动驾驶汽车障碍物识别判据量化的方法。通过激光雷达完成障碍物点云信息的采集,采用坐标变换、XYZ三向平面切分、平面内椭圆曲线的建立及判据量化计算等一系列数学方法实现障碍物识别判据的量化,为自动驾驶汽车在行使过程中关于障碍物是否侵犯汽车驾驶安全空间提供了一种可量化的判断准则。
In order to realize the quantification of obstacle recognition criteria for autonomous vehicles,this paper proposes a method for the quantification of obstacle recognition criteria for autonomous vehicles based on elliptic envelope curve sets.The collection of obstacle point cloud information is completed by lidar,and then a series of mathematical methods such as coordinate transformation,XYZ three-way plane segmentation,establishment of elliptic curves in the plane,and criterion quantification calculation are used to realize the quantification of obstacle identification criteria.The autonomous vehicle provides a quantifiable criterion for determining whether obstacles infringe the safe driving space of the vehicle during the driving process.
作者
高少波
胡瑜
张笑
GAO Shao-bo;HU Yu;ZHANG Xiao(Zhejiang Lab Advanced Computer Research Center,Hangzhou 310000,China;Institute of Computing Technology,Chinese Academy of Sciences,Beijing 100190,China)
出处
《信息技术》
2024年第3期83-89,97,共8页
Information Technology
基金
之江实验室科研攻关项目(2022PC0AC01)。
关键词
自动驾驶
椭圆
包络
判据
量化
automatic driving
elliptic
envelope
criterion
quantization
