基于图像感兴趣区域的机器人闭环检测算法

doi:10.13973/j.cnki.robot.180618

摘要
图/表
参考文献
相关文章 (15)

全文: PDF (489 KB) HTML ( ) 输出: BibTeX \| EndNote (RIS)
引用本文:
潘锡英, 何元烈, 孙盛, 陈佳腾. 基于图像感兴趣区域的机器人闭环检测算法[J]. 机器人, 2019, 41(5): 676-682.DOI: 10.13973/j.cnki.robot.180618. PAN Xiying, HE Yuanlie, SUN Sheng, CHEN Jiateng. A Loop Closure Detection Algorithm for Robots Based on Region Proposals of Interest of Image. ROBOT, 2019, 41(5): 676-682. DOI: 10.13973/j.cnki.robot.180618.

摘要基于深度学习的机器人闭环检测算法在复杂光照条件下具有一定的鲁棒性，但在视角变化明显的场景下检测效果不佳，为此本文提出一种利用图像感兴趣区域的闭环检测新方法．首先，通过多尺度感兴趣区域网络（MSRPN）获得图像中的感兴趣区域，用改进的PlaceCNN（基于Place数据集的卷积神经网络）提取感兴趣区域的特征．然后，采用先粗匹配后细匹配原则，提出一种基于RPOI_PlaceCNN（基于图像感兴趣区域的PlaceCNN）的闭环检测算法，并利用双向匹配对之间的空间约束，去除不正确的匹配对，以提高闭环检测的整体准确性．在GardensPoint、Mapillary、Norland三种公开数据集上对方法的有效性进行了实验验证．实验结果表明，本文提出的闭环检测算法在光照、视角和不同变化组合引起的显著变化场景下依然能表现出较强的鲁棒性．

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	潘锡英
	何元烈
	孙盛
	陈佳腾

关键词 ：闭环检测, 卷积神经网络, 特征提取, 场景识别, 特征匹配, 视觉同步定位与地图构建

Abstract：The loop closure detection algorithm of robots based on deep learning shows certain robustness under complex illumination, but it is prone to be affected by the scene with obvious changes of view angle. Therefore, a new loop closure detection method using RPOIs (region proposals of interest) of image is proposed. Firstly, the RPOIs of image are obtained by the improved MSRPN (multi-scale region proposal network), and the feature of interested region proposals is extracted by the improved PlaceCNN (Place dataset based convolutional neural network). Then, considering the shape similarity of region proposals, a loop closure detection algorithm based on RPOI_PlaceCNN (RPOI based PlaceCNN) is proposed by adopting the principle of coarse matching firstly and fine matching secondly. The space constraint between bidirectional matching pairs is used to remove incorrect matching pairs, which can improve the overall accuracy of loop closure detection. The effectiveness of the proposed method is experimentally verified on three public datasets, i.e. GardensPoint, Mapillary and Norland datasets. The experimental results show that the loop closure detection algorithm proposed can exhibit strong robustness in the situations of significant scene changes caused by illumination, view angle and different combinations of changes.

Key words： loop closure detection convolutional neural network feature extraction scene recognition feature matching visual simultaneous localization and mapping

收稿日期: 2018-10-16 录用日期: 2019-02-28

TP24

基金资助:国家自然科学基金（41501362）

通讯作者: 何元烈,heyuanlie@163.com E-mail: heyuanlie@163.com

作者简介: 潘锡英(1992-),女,硕士生.研究领域:计算机视觉,图像处理,深度学习.
何元烈(1976-),男,博士,副教授.研究领域:计算机视觉,深度学习,智能机器人.
孙盛(1980-),男,博士,副教授.研究领域:遥感图像处理,计算机视觉.

链接本文:

https://robot.sia.cn/CN/10.13973/j.cnki.robot.180618 或 https://robot.sia.cn/CN/Y2019/V41/I5/676

[1] 刘浩敏,章国锋,鲍虎军.基于单目视觉的同时定位与地图构建方法综述[J].计算机辅助设计与图形学学报,2016,28(6):855-868.Liu H M, Zhang G F, Bao H J. A survey of monocular simultaneous localization and mapping[J]. Journal of Computer-Aided Design and Computer Graphics, 2016, 28(6):855-868.
[2] 梁明杰,闵华清,罗荣华.基于图优化的同时定位与地图创建综述[J].机器人,2013,35(4):500-512.Liang M J, Ming H Q, Luo R H. Graph-based SLAM:A survey[J]. Robot, 2013, 35(4):500-512.
[3] Lowe D G. Distinctive image features from scale-invariant keypoints[J]. International Journal of Computer Vision, 2004, 60(2):91-110.
[4] 包加桐,宋爱国,郭晏,等.基于SURF特征跟踪的动态手势识别算法[J].机器人,2011,33(4):482-489.Bao J T, Song A G, Guo Y, et al. Dynamic gesture recognition algorithm based on SURF feature tracking[J]. Robot, 2011, 33(4):482-489.
[5] Mur-Artal R, Tardos J D. ORB-SLAM2:An open-sourceSLAM system for monocular, stereo, and RGB-D cameras[J]. IEEE Transactions on Robotics, 2017, 33(5):1255-1262.
[6] Sivic J, Zisserman A. Video Google:A text retrieval approach to object matching in videos[C]//IEEE International Conference on Computer Vision. Piscataway, USA:IEEE, 2003:1470-1477.
[7] Liu Y, Zhang H. Visual loop closure detection with a compact image descriptor[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems. Piscataway, USA:IEEE, 2012:1051-1056.
[8] 赵洋,刘国良,田国会,等.基于深度学习的视觉SLAM综述[J].机器人,2017,39(6):889-896. Zhao Y, Liu G L, Tian G H, et al. A survey of visual SLAM based on deep learning[J]. Robot, 2017, 39(6):889-896.
[9] Krizhevsky, A, Sutskever, I, Hinton, G E. ImageNet classification with deep convolutional neural networks[C]//Advances in Neural Information Processing Systems. San Diego, USA:Neural Information Processing Systems Foundation, 2012:1097-1105.
[10] Ren S Q, He K M, Girshick R, et al. Faster R-CNN:Towards real-time object detection with region proposal networks[C]//Advances in Neural Information Processing Systems. San Diego, USA:Neural Information Processing Systems Foundation, 2015:91-99.
[11] Hosang J, Benenson R, Dollár P, et al. What makes for effective detection proposals?[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 38(4):814-830.
[12] Cai Z W, Fan Q F, Feris R S, et al. A unified multi-scale deep convolutional neural network for fast object detection[C]//14th European Conference on Computer Vision. Cham, Switzerland:Springer, 2016:354-370.
[13] Gao X, Zhang T. Unsupervised learning to detect loops using deep neural networks for visual SLAM system[J]. Autonomous Robots, 2017, 41(1):1-18.
[14] Sünderhauf N, Shirazi S, Dayoub F, et al. On the performance of ConvNet features for place recognition[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems. Piscataway, USA:IEEE, 2015:4297-4304.
[15] Zhou B, Lapedriza A, Khosla A, et al. Places:A 10 million image database for scene recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(6):1452-1464.
[16] Hou Y, Zhang H, Zhou S L. Convolutional neural network-based image representation for visual loop closure detection[C]//IEEE International Conference on Information and Automation. Piscataway, USA:IEEE, 2015:2238-2245.
[17] Kong T, Yao A B, Chen Y R, et al. HyperNet:Towards accurate region proposal generation and joint object detection[C]//IEEE Conference on Computer Vision and Pattern Recognition. Piscataway, USA:IEEE, 2016:845-853.
[18] He K M, Zhang X Y, Ren S Q, et al. Delving deep into rectifiers:Surpassing human-level performance on ImageNet classification[C]//IEEE International Conference on Computer Vision. Piscataway, USA:IEEE, 2015:1026-1034.
[19] Mapillary. Mapillary[EB/OL].[2016-05-15]. https://www.mapillary.com.
[20] Neubert P, Sünderhauf N, Protzel P. Superpixel-based appearance change prediction for long-term navigation across seasons[J]. Robotics and Autonomous Systems, 2015, 69(S1):15-27.