公开道路车道线真值测试评价方法研究

Research on The Evaluation Method of The Truth Value Test of Public Road Lane Lines

随着L2级ADAS功能的普及,车辆需要识别车道线来进行横向控制。公开道路测试是智能驾驶车辆测试“三支柱”的最后环节,通过公开道路测试的方法来验证车道线的识别效果必不可少。本文就当前公开道路车道线测试方法的欠缺,提出多传感器数据层及特征层融合的方案,利用车道线摄像头辅助激光雷达的方案做为车道线测试冗余的真值系统,并提出相应的数据分析指标,最后通过实车测试验证了该方案的有效性。

With the popularity of L2-level ADAS functions,vehicles need to identify lane lines for lateral control.The public road test is the last link of the“three pillars”of the intelligent driving vehicle test.It is essential to verify the recognition effect of the lane line by the method of the public road test.In view of the lack of current public road lane line testing methods,this paper proposes a multi-sensor data layer and feature layer fusion scheme,using the lane line camera-assisted lidar scheme as a redundant true value system for lane line testing,and proposes corresponding data The indicators are analyzed,and finally the effectiveness of the scheme is verified through the real vehicle test.