基于激光雷达数据SVM分类的室内环境识别研究
PDF下载 (363)郭玉朋,刘士荣,张波涛.基于激光雷达数据SVM分类的室内环境识别研究[J].宁波大学学报(理工版),2013,26(04):35-39.DOI:
GUO Yu-peng,LIU Shi-rong,ZHANG Bo-tao.Indoor Environment Recognition Based on SVM Classification of Laser Radar Data[J].Journal of Ningbo University(Natural Science & Engineering Edition),2013,26(04):35-39.DOI:
| Title: | Indoor Environment Recognition Based on SVM Classification of Laser Radar Data |
| 作者: | 郭玉朋, 刘士荣, 张波涛 |
| Author(s): | GUO Yu-peng, LIU Shi-rong, ZHANG Bo-tao |
| 关键词: | 室内环境; 语义地图; SVM分类器; 移动机器人 |
| Keywords: | indoor environment; semantic map; SVM classifier; mobile robot |
| 分类号: | TP24 |
| 文献标识码: | A |
| 摘要: | 语义地图构建对移动机器人导航与规划具有重要意义, 而环境分类是语义地图构建的核心问题. 目前所采用的环境分类方法匹配率较低, 已成为语义地图构建所面临的主要问题. 对此笔者提出了一种基于支持向量机的分类方法, 该方法利用激光雷达数据提取环境几何特征, 训练SVM分类器对机器人工作空间模式进行识别, 并将所提算法用于室内环境的语义分类. 实验结果表明, 该分类方法具有较高的识别率, 可有效地实现语义地图构建. |
| Abstract: | Semantic map building is of significant importance for the navigation and trajectory planning of mobile robots, in which the environmental classification has been a key issue to address. The lower matching rate has been identified in the currently adopted environmental classification methods, which forms a technical bottle-neck for the semantic map building. A classification method based on support vector machine (SVM) is proposed in this paper, In the process, the geometric features extracted from the laser radar data are employed to train the SVM classifier, which is used for identifying he robotic workspace pattern. The proposed method is applied to the semantic classification of the given indoor layout. The experiment results suggest that the proposed method is capable of building the semantic map efficiently with a higher recognition rate. |
| 参考文献 /References: | [1] 蔡自兴, 贺汉根, 陈虹.未知环境中移动机器人导航控制理论与方法[M]. 北京: 科学出版社, 2009. [2] Luo R C, Lai C C. Concurrent indoor map construction and patterns of interests recognition using sensory fusion approach for service robotics[C]//IEEE International Conference on Robotics and Automation (ICRA). St. Paul: IEEE, 2012:2243-2248. [3] Elfes A. Using occupancy grids for mobile robot perception and navigation[J]. Computer, 1989, 22(6):46- 57. [4] Leonard J, Durrant-Whyte H. Dynamic map building for an autonomous mobile robot[J]. The International Journal on Robotics Research, 1992, 11(4):286-298. [5] Thrun S, Gutmann J S, Fox D, et al. Integrating topological and metric maps for mobile robot navigation: A statistical approach[C]//Proceedings of the National Conference on Artificial Intelligence. Madison: JOHN WILEY & SONS LTD, 1998:989-996. [6] 吴皓.基于服务任务导向的移动机器人地图构建研究[D]. 济南: 山东大学, 2011. [7] Ila V, Porta J M, Andrade-Cetto J. Information-based compact pose SLAM[J]. IEEE Transactions on Robotics, 2010, 26(1):78-93. [8] Mozos ? M. Semantic labeling of places with mobile robots[M]. Berlin: Springer Verlag, 2010. [9] 丁刚. 基于支持向量机的移动机器人环境感知和物体识别研究[D]. 合肥: 中国科技大学, 2009. |
| 备注/Memo: | 收稿日期: 2013-05-04. 基金项目: 国家自然科学基金（61175093）. 第一作者: 郭玉朋（1987-）, 男, 山东郓城人, 在读硕士研究生, 主要研究方向: 机器人与智能系统. E-mail: guoyupeng11@163.com *通信作者: 刘士荣（1952-）, 男, 浙江绍兴人, 博士/教授, 主要研究方向: 智能机器人与智能控制等. E-mail: liushirong@hdu.edu.cn 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ |
