无人驾驶车前方不确定障碍物检测方法仿真.pdfVIP

第40卷第2期 计 算 机 仿 真 2023年2月 文章编号：1006-9348(2023)02-0471-04 无人驾驶车前方不确定障碍物检测方法仿真 田国红，关亮亮 （辽宁工业大学汽车与交通工程学院，辽宁锦州121001） 摘要：无人驾驶车障碍物检测是保障车辆安全运行的关键，但是障碍物检测过程中环境对检测方法的影响较大，为提升无人 驾驶车在不确定环境中的障碍物检测精度，提出不确定环境下无人驾驶车前方障碍物检测方法。利用激光雷达获取汽车行 驶过程的激光雷达图像，通过双边滤波方法完成图像的去噪处理，依据搜索的道路边界点建立道路边界模型，提取障碍物特 征建立障碍物模型；通过时空特征系数分类结果，完成障碍物类型识别，实现障碍物检测。实验结果表明，所提方法具有较 低的误检率和漏检率，能够精准地检测出前方障碍物位置。 关键词：激光雷达；无人驾驶车：前方障碍物；数据预处理；双边滤波 中图分类号：TP391 文献标识码：B Simulation of Detection Method for Uncertain Obstacles in Front of Unmanned Vehicles TIAN Guo-hong,GUAN Liang-liang ( School of Automobile and Traffic Engineering, Liaoning University of Technology ,Jinzhou Liaoning 121001, China) ABSTRACT: At present,the detection of obstacles in front of driverless cars is the key to ensure the safe operation of vehicles,but the environment has a big effect on the detection method. In order to improve the detection accuracy in an uncertain environment,this paper presented a method of detecting obstacles in front of driverless cars. At first,the lidar image of the car in the course of driving was obtained by lidar, and then image denoising was completed by the bilateral filtering method. Moreover ,a road boundary model was built based on the searched boundary points. Mean- while,the obstacle features were extracted while building the obstacle model. Finally, based on the classification results of spatio-temporal characteristic coefficients, the recognition for obstacle type was completed. Thus, obstacle detection was achieved. Experimental results show that the proposed method has a low false detection rate and missed detection rate , and can detect the position of the obstacles in front