Skip to content

gilbertgonz/visual_slam

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Monocular Visual SLAM

Welcome to my VSLAM project! This project aims to track the single camera's motion and estimate its pose using ORB feature detection (also tested using SIFT, but way too computationally expensive...), triangulate 3D points from consecutive frames, and optimize the estimated camera poses and 3D points through bundle adjustment (BA). The resulting trajectory and point cloud are visualized using Pangolin. As can be seen in the gif below, I am utilizing short snips from the well-known KITTI dataset.

To run:

  1. Install docker

  2. Clone repo

  3. Build:

$ docker build -t visual_odom .
  1. Run:
$ xhost +local:docker
$ docker run --rm -e DISPLAY=$DISPLAY -v /tmp/.X11-unix:/tmp/.X11-unix visual_odom

Future work:

This project is still on-going, things I still plan to implement/test:

  • Test performance on aerial robot (drone) image data
  • Implement a dedicated backend thread for proper execution of BA
  • Deploy on hardware platform and implement real-time operation

About

Monocular SLAM implemetation with KITTI dataset

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published