首页 > 论文 > 电光与控制 > 25卷 > 12期(pp:30-34)

基于2-D范围扫描的室内场景识别方法

An Indoor Scene Recognition Method Based on 2-D Range Scanning

  • 摘要
  • 论文信息
  • 参考文献
  • 被引情况
  • PDF全文
分享:

摘要

对于在室内工作的机器人而言,要完成不同环境的自主导航就必须能够对其所处场景进行有效识别。传统的方法是通过视觉或雷达等传感器对所处环境进行匹配以实现场景识别。提出了一种基于2-D范围扫描的室内场景识别的方法。该方法对激光雷达的范围扫描信息进行特征提取,利用所提取的样本训练基于局部感受野的极限学习机,对多种室内场景进行分类识别。在Gazebo搭建的仿真环境中采集虚拟范围扫描数据,对室内场景识别方法进行了研究。利用DR Dataset提供的测距数据对所提出的方法进行了实验验证。实验结果表明: 该方法的室内场景识别准确率高于传统方法。基于2-D范围扫描场景识别的研究也为机器人实现自主导航提供理论依据和实验数据。

Abstract

The robots working indoors must be able to effectively identify their surroundings to complete the autonomous navigation in different scenes.Traditional approaches realize scene recognition by using visual or radar sensors to match the scene.A method of indoor scene recognition based on 2-D range scanning is proposed.This method extracts the features of range scanning information of the lidar, and Extreme Learning Machine Based on Local Receptive Fields (ELM-LRF) is trained by using extracted samples to classify and identify various indoor scenes.In the simulation environment built by Gazebo, the virtual range scanning data is collected, and then the indoor scene recognition methods are studied.The proposed method is verified by experiments based on the range data provided by DR Dataset.The results show that the recognition accuracy of the proposed method is higher than that of traditional methods.The study of scene recognition based on 2-D range scanning also provides theoretical support and experimental data for autonomous robot navigation.

Newport宣传-MKS新实验室计划
补充资料

中图分类号:TP181

DOI:10.3969/j.issn.1671-637x.2018.12.007

所属栏目:学术研究

基金项目:国家自然科学基金(51777053); 河北省自然科学基金(E2017202035);河北省高层次人才项目 (C2015003037)

收稿日期:2017-12-19

修改稿日期:2018-02-12

网络出版日期:--

作者单位    点击查看

郜春艳:河北工业大学省部共建电工装备可靠性与智能化国家重点实验室,天津 300132
何秀娟:河北工业大学省部共建电工装备可靠性与智能化国家重点实验室,天津 300132
黄文美:河北工业大学省部共建电工装备可靠性与智能化国家重点实验室,天津 300132
刘卓锟:河北工业大学省部共建电工装备可靠性与智能化国家重点实验室,天津 300132

联系人作者:联系作者

备注:郜春艳(1991—),女,江苏盐城人,硕士,研究方向为机器学习和模式识别、新型磁性材料与器件。

【1】ORABONA F, CASTELLINI C, CAPUTO B, et al.Indoor place recognition using online independent support vector machines[C]//Proceedings of British Machine Vision Conference, 2007: 1090-1099.

【2】许坤,邹杰,陈谋.基于联邦滤波的室内移动机器人定位[J].电光与控制,2016,23(9): 77-83, 110.

【3】URSIC P, LEONARDIS A, SKOCAJ D, et al.Hierarchical spatial model for 2D range data based room categorization[C]//IEEE International Conference on Robotics and Automation, 2016: 4514-4521.

【4】LIU H P, SUN F C, FANG B, et al.Robotic room-level localization using multiple sets of sonar measurements[J].IEEE Transactions on Instrumentation and Measurement, 2016, 66(1): 2-13.

【5】PARK S, ROH K S.Coarse-to-fine localization for a mobile robot based on place learning with a 2-D range scan[J].IEEE Transactions on Robotics, 2016, 32(3): 528-544.

【6】DEMARCO K J, WEST M E, HOWARD A M.A computationally-efficient 2D imaging sonar model for underwater robotics simulations in Gazebo[C]//OCEANS, 2015: 1-7.

【7】TAKAYA K, ASAI T, KROUMOV V, et al.Simulation environment for mobile robots testing using ROS and Gazebo[C]//The 20th International Conference on System Theory, Control and Computing (ICSTCC), 2016: 96-101.

【8】QIAN W, XIA Z Y, XIONG J, et al.Manipulation task simulation using ROS and Gazebo[C]//IEEE International Conference on Robotics and Biomimetics(ROBIO), 2014: 2594-2598.

【9】HUANG G B, BAI Z, KASUN L L C.Local receptive fields based extreme learning machine[J].IEEE Computational Intelligence Magazine,2015, 10(2): 18-29.

【10】LYU Q, NIU X, DOU Y, et al.Hyperspectral image classification via kernel extreme learning machine using local receptive fields[C]//IEEE International Conference on Image Processing (ICIP), 2016: 256-260.

【11】李桂芝,安成万,杨国胜,等.基于场景识别的移动机器人定位方法研究[J].机器人,2005, 27(2): 123-127.

【12】鄢文浩,贺赛先,沈婷婷.基于随机运动障碍避碰规则的机器人路径规划[J].电光与控制,2017, 24(5): 73-76.

引用该论文

GAO Chun-yan,HE Xiu-juan,HUANG Wen-mei,LIU Zhuo-kun. An Indoor Scene Recognition Method Based on 2-D Range Scanning[J]. Electronics Optics & Control, 2018, 25(12): 30-34

郜春艳,何秀娟,黄文美,刘卓锟. 基于2-D范围扫描的室内场景识别方法[J]. 电光与控制, 2018, 25(12): 30-34

您的浏览器不支持PDF插件,请使用最新的(Chrome/Fire Fox等)浏览器.或者您还可以点击此处下载该论文PDF