Canny-EVT is a C++/ROS software framework for semi-dense 6-DoF tracking with event cameras. It accompanies the paper submission "Semi-Dense Cross-Modal 6-DoF Tracking of an Event Camera in Challenging Conditions" and provides documented modules for event-camera modeling, time-surface construction, pose optimization, and dataset evaluation.

Related Publications:

Zuo, Yi-Fan, et al. "DEVO: Depth-Event Camera Visual Odometry in Challenging Conditions." 2022 International Conference on Robotics and Automation (ICRA). IEEE, 2022.
He, Shida, et al. "Incremental 3d line segment extraction from semi-dense slam." 2018 24th International Conference on Pattern Recognition (ICPR). IEEE, 2018.
Mur-Artal, Raúl, and Juan D. Tardós. "Probabilistic Semi-Dense Mapping from Highly Accurate Feature-Based Monocular SLAM." Robotics: Science and Systems. Vol. 2015. 2015.

Notice

Please be aware that this open-source code is still in the debugging and testing phase of development. It may not be stable and is likely subject to change.

If you are seeking to replicate the test results as presented in the associated paper, we have provided an executable version for your convenience. For detailed information and access to the executable file, please visit our repository:

Canny-VOAT Executable File

Getting started

Canny-EVT includes the following core modules:

Event-camera modeling and event data handling.
Time-surface construction from event streams.
LM-based semi-dense 6-DoF pose optimization.
ROS-based running, dataset conversion, and evaluation utilities.

The aim of Canny-EVT is to make the implementation and experiments for semi-dense event-camera tracking easier to run, inspect, and reproduce on the supported datasets.

Please consult the following sub-pages to get a step-by-step introduction to the library: