欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
车载激光雷达所获取的程距数据有数据量大、数据流有序、噪声点多等特点,文中针对激光雷达数据的特点提出了一种带约束的自适应聚类分析方法。该方法通过人机交互输入参数以后,对激光雷达的扫描点进行自动聚类,通过迭代后得到相对光滑的线段,为道路障碍检测等后续工作提供了便利。最后还讨论了如何通过计算线段的位置关系,得到道路方向、障碍位置及大小等信息。通过对程距数据的实验表明,该方法能够达到预期的效果。
英文摘要：
The range data of lidar on vehicle have many characters such as large amount, ordered data and with much noises. In view of the characters of range data, a method of clustering analysis with restriction is presented in this paper. In the proposed method, the scan points have been automatically clustered after parameters are input from man - machine conversation, while relatively smooth line - segments can be obtained after iterative computing. It provides convenience for later works such as road and obstacle detecting. At the end of this paper, how to get the position and size of the road and obstacles by computing the relationship of the line - segments is discussed. The experimental results on range data show that this clustering algorithm can attain expected performance.
同期刊论文项目
同项目期刊论文
期刊信息
位置：