Fast loop-closure detection using visual-word-vectors from image sequences

International Journal of Robotics Research - Tập 37 Số 1 - Trang 62-82 - 2018
Loukas Bampis1, Angelos Amanatiadis1, Αντώνιος Γαστεράτος1
1Department of Production and Management Engineering, Democritus University of Thrace, Greece

Tóm tắt

In this paper, a novel pipeline for loop-closure detection is proposed. We base our work on a bag of binary feature words and we produce a description vector capable of characterizing a physical scene as a whole. Instead of relying on single camera measurements, the robot’s trajectory is dynamically segmented into image sequences according to its content. The visual word occurrences from each sequence are then combined to create sequence-visual-word-vectors and provide additional information to the matching functionality. In this way, scenes with considerable visual differences are firstly discarded, while the respective image-to-image associations are provided subsequently. With the purpose of further enhancing the system’s performance, a novel temporal consistency filter (trained offline) is also introduced to advance matches that persist over time. Evaluation results prove that the presented method compares favorably with other state-of-the-art techniques, while our algorithm is tested on a tablet device, verifying the computational efficiency of the approach.

Từ khóa


Tài liệu tham khảo

10.1109/IROS.2015.7353807

10.1109/TRO.2008.2004514

10.1109/ICRA.2015.7140088

10.1109/IROS.2016.7759685

Arthur D, 2007, Proceedings of the eighteenth annual ACM-SIAM symposium on discrete algorithms, 1027

10.1109/IROS.2016.7759667

10.1007/11744023_32

10.1007/s10514-009-9138-7

Chandrasekhar V, 2014, Proceedings of the IEEE conference on computer vision and pattern recognition workshops, 527

10.1109/ICPR.2014.454

10.1109/LRA.2017.2650153

10.1016/j.ijforecast.2011.07.006

10.1177/0278364908090961

10.1177/0278364910385483

10.1109/TPAMI.2007.1049

10.1109/ICRA.2015.7139962

10.1007/978-3-319-46487-9_20

10.1109/ROBOT.2004.1307180

10.1109/IROS.2007.4399123

10.1109/TRO.2012.2197158

10.1177/0278364913491297

Google (2017) Project Tango. Available at: https://developers.google.com/tango/hardware/tablet (accessed 10 October 2017).

10.1109/MITS.2010.939925

10.1007/978-3-540-88682-2_24

10.1109/ICCV.2011.6126328

10.1109/ICRA.2015.7139959

10.1093/oxfordjournals.pan.a004868

10.1109/ISMAR.2007.4538852

10.1177/0278364910370376

10.1109/TRO.2013.2242375

10.1177/0278364913498910

Lee JH, 2013, Proceedings of the IEEE international conference on robotics and automation, 3799

10.1109/ICRA.2014.6907055

10.1109/ICRA.2016.7487208

10.1023/B:VISI.0000029664.99615.94

10.1109/TRO.2015.2496823

10.1109/3DV.2014.36

10.1007/s10514-015-9463-y

MacTavish K, 2014, Proceedings of the IEEE international conference on robotics and automation, visual place recognition in changing environments workshop

Maddern W, 2014, Proceedings of the IEEE international conference on robotics and automation, visual place recognition changing environments workshop

10.5244/C.23.54

10.1109/ICRA.2012.6224623

10.1109/IROS.2016.7759098

10.1109/ICRA.2014.6906953

10.1109/TRO.2015.2463671

10.1109/ROBOT.2006.1641869

10.1109/CVPRW.2015.7301272

10.1109/CVPR.2006.264

10.1002/rob.20355

Rawseeds (2007) The Rawseeds Project. Available at: http://www.rawseeds.org/rs/datasets (accessed 10 October 2017).

10.1007/11744023_34

10.1109/ICCV.2011.6126544

10.1109/CVPR.2007.383150

10.1109/IROS.2016.7759095

10.1109/ICCV.2003.1238663

10.1007/978-3-319-49409-8_74

10.1177/0278364909103911

10.15607/RSS.2010.VI.010

10.1109/IROS.2015.7353986

10.15607/RSS.2015.XI.022

10.1177/0278364906065387

10.1016/j.robot.2009.09.010

10.1016/j.robot.2009.06.010

10.1109/ICASSP.1984.1172729

10.1109/TRO.2004.835453

Thông tin
Thông tin xuất bản

International Journal of Robotics Research

Tập 37 Số 1

62-82

Thông tin tác giả