基于四叉树扇形层值聚类的无人船障碍物检测
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

TP273

基金项目:

工信部高技术船舶科研项目([2019]360号);张家港市科技计划项目(ZKC2206; ZKYY2253)


Obstacle Detection for Unmanned Ship Based on Quadtree Sector Layer Value Clustering
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    为了实现无人船自主导航过程中对障碍物的精确检测,提出了一种基于四叉树扇形层值聚类的无人船障碍物检测方法。首先基于四叉树扇形划分进行障碍物点云数据地检索,并剔除扇形象限内不可信数据;然后利用所获得的四叉树层值来求取全局密度距离,进而获得层值阈值,以此来对不规则多线形障碍物特征进行检测;最后通过建立数据点之间的空间拓扑关系来求取参考距离,并以参考距离为基准对障碍物点云数据进行聚类判定,提高聚类分割准确性。多线形障碍物特征识别性能测试及水面无人船障碍物检测实验结果表明,相较于其他密度聚类算法,在正检率、误检率和漏检率性能指标方面,前者测试中文中算法分别平均下降了9.86%、5.04%、3.10%,后者测试中文中算法分别平均下降了10.50% 、6.97%、2.95%。

    Abstract:

    In order to realize the precise detection of obstacles in the process of autonomous navigation of unmanned ships, an obstacle detection method for unmanned ships based on quadtree sector layer value clustering is proposed. Firstly, the obstacle point cloud data is retrieved based on the quadtree sector division, and the untrustworthy data in the sector image limit is eliminated; then the obtained quadtree layer value is used to calculate the global density distance, and then the layer value threshold is obtained. to detect irregular multi-linear obstacle features; finally, the reference distance is obtained by establishing the spatial topological relationship between data points, and the obstacle point cloud data is clustered and judged based on the reference distance to improve cluster segmentation accuracy. The multi-linear obstacle feature recognition performance test and the surface unmanned ship obstacle detection experiment results show that, compared with other density clustering algorithms, in terms of positive detection rate, false detection rate and missed detection rate performance indicators, the former test algorithm in Chinese The average decreases were 9.86%, 5.04%, and 3.10%, respectively, and the algorithms in the latter test Chinese decreased by an average of 10.50%, 6.97%, and 2.95%, respectively.

    参考文献
    相似文献
    引证文献
引用本文

申燚,赵泽钰,袁明新,等. 基于四叉树扇形层值聚类的无人船障碍物检测[J]. 科学技术与工程, 2024, 24(13): 5427-5435.
Shen Yi, Zhao Zeyu, Yuan Mingxin, et al. Obstacle Detection for Unmanned Ship Based on Quadtree Sector Layer Value Clustering[J]. Science Technology and Engineering,2024,24(13):5427-5435.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2023-06-01
  • 最后修改日期:2023-09-24
  • 录用日期:2023-09-30
  • 在线发布日期: 2024-05-17
  • 出版日期:
×
亟待确认版面费归属稿件,敬请作者关注
《科学技术与工程》入选维普《中文科技期刊数据库》自然科学类期刊月度下载排行榜TOP10