Skip to content

A Feature-based visual SLAM algorithm which takes video footage as input

Notifications You must be signed in to change notification settings

Brad1141/vidSLAM

Repository files navigation

vidSLAM

What is vidSLAM?

vidSLAM is a Feature-based visual SLAM program which preforms SLAM on a given video using Structure from Motion(SfM)

Installing and Running vidSLAM

  1. Install the dependencies listed below
  2. run the command git clone https://github.com/Brad1141/vidSLAM.git
  3. Open the repo and run the filemain.py
  • main.py accepts two arguments -video and -fov
  • -video The path of the video file (the default video is drive.mp4)
  • -fov The field of view for the camera of the video (The default is 60)

Dependancies

  • Matplotlib pip install matplotlib
  • OpenCV pip install opencv-python
  • Numpy pip install numpy

More Info

This project was originally part of a tutorial I made on how to do feature based visual SLAM. Feel free to check out the tutorial series for more details on the methods and features used in vidSLAM.

About

A Feature-based visual SLAM algorithm which takes video footage as input

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages