Ben Talbot
26 March 2017
Last updated
4 February 2021

1 August 2017

First planning meeting for OpenSeqSLAM2 project

8 December 2017

Alpha release, shared internally in QUT's robotics group (v0.1.0)

26 March 2017

Stable version released, used internally for comparative visual place recognition experiments (v1.0.0)

6 April 2017

Released publicly for the first time (v2.0.0)

1 October 2018

Toolbox published in proceedings of IROS 2018

4 February 2021

Migrated to GitHub, and added to QCR's open source portfolio (v2.0.1)

Snapshots of the OpenSeqSLAM2 user interface

OpenSeqSLAM2.0 is a MATLAB toolbox that allows users to thoroughly explored the SeqSLAM method for addressing the visual place recognition problem. The visual place recognition problem is centred around recognising a previously traversed route, regardless of whether it is seen during the day or night, in clear or inclement conditions, or in summer or winter. Recognising previously traversed routes is a crucial capability of navigating robots.

The toolbox provides a number of easy-to-use graphical interfaces that allow users to interactively learn about the SeqSLAM algorithm by exploring and playing with its underlying behaviour. Through visual GUIs users are able to:

  • explore a number of previously published variations to the SeqSLAM method (including search and match selection methods);
  • visually track progress;
  • interactively tune parameters;
  • dynamically reconfigure matching parameters while viewing results;
  • explore precision-recall statistics;
  • visualise difference matrices, match sequence images, and image pre-processing steps;
  • view and export matching videos;
  • automatically optimise selection thresholds against a ground truth;
  • sweep any numeric parameter value through a batch operation mode; and
  • operate in headless mode with parallelisation available.

The toolbox is open-source and available on GitHub. See the links below for further details.


Sourav GargMichael Milford

Related Links

View the code on GitHubFull details of 2018 IROS paper

© Ben Talbot. All rights reserved.