基于单线激光雷达与视觉融合的负障碍检测算法
发布时间:2018-08-10 17:17
【摘要】:近年来,无人车成为热门研究方向,而负障碍物检测是地面无人车环境感知与理解的任务之一。为此,提出一种基于单线激光雷达和单目视觉的负障碍检测算法。为弥补单线激光雷达在覆盖能力方面的不足,对检测到的负障碍区域在摄像机画面中进行跟踪,结合跟踪结果对负障碍区域做进一步判别。实验结果表明,该算法在多种实验场景下拥有96%以上的负障碍检测准确率,可有效应用于微小型地面无人车辆。
[Abstract]:In recent years, unmanned vehicle (UAV) has become a hot research direction, and negative obstacle detection is one of the tasks of environment perception and understanding of ground UAV. Therefore, a negative obstacle detection algorithm based on single line lidar and monocular vision is proposed. In order to make up for the lack of coverage ability of single line lidar, the detected negative obstacle area is tracked in the camera picture, and the negative obstacle area is further judged by the tracking results. The experimental results show that the algorithm has more than 96% accuracy of negative obstacle detection in many experimental scenarios, and it can be effectively applied to small unmanned vehicles on the ground.
【作者单位】: 南京理工大学计算机科学与工程学院;
【基金】:国家自然科学基金(61272220) 江苏省自然科学基金(BK20140794)
【分类号】:TN958.98
[Abstract]:In recent years, unmanned vehicle (UAV) has become a hot research direction, and negative obstacle detection is one of the tasks of environment perception and understanding of ground UAV. Therefore, a negative obstacle detection algorithm based on single line lidar and monocular vision is proposed. In order to make up for the lack of coverage ability of single line lidar, the detected negative obstacle area is tracked in the camera picture, and the negative obstacle area is further judged by the tracking results. The experimental results show that the algorithm has more than 96% accuracy of negative obstacle detection in many experimental scenarios, and it can be effectively applied to small unmanned vehicles on the ground.
【作者单位】: 南京理工大学计算机科学与工程学院;
【基金】:国家自然科学基金(61272220) 江苏省自然科学基金(BK20140794)
【分类号】:TN958.98
【参考文献】
相关期刊论文 前5条
1 王盛;项志宇;;基于多谱融合的植被环境中障碍物检测[J];浙江大学学报(工学版);2015年11期
2 张建勋;汪波;侯之旭;靳冲;;图像多特征融合的障碍物检测[J];重庆理工大学学报(自然科学);2015年03期
3 彭梦;蔡自兴;;一种基于双平行平面的激光雷达和摄像机标定方法[J];中南大学学报(自然科学版);2012年12期
4 刘大学;戴斌;李政;贺汉根;;一种单线激光雷达和可见光摄像机的标定方法[J];华中科技大学学报(自然科学版);2008年S1期
5 张宏志;张金换;岳卉;黄世霖;;基于CamShift的目标跟踪算法[J];计算机工程与设计;2006年11期
相关硕士学位论文 前1条
1 史鹏波;基于单线激光雷达的道路特征检测[D];南京理工大学;2013年
【共引文献】
相关期刊论文 前10条
1 汪佩;郭剑辉;李伦波;赵春霞;;基于单线激光雷达与视觉融合的负障碍检测算法[J];计算机工程;2017年07期
2 贾子永;任国全;李冬伟;程子阳;;基于梯形棋盘格的摄像机和激光雷达标定方法[J];计算机应用;2017年07期
3 杨幸;王力;;基于OpenCV的视频人脸检测与跟踪[J];通信技术;2017年07期
4 董凌翔;林巍\,
本文编号:2175628
本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/2175628.html