Skip to content

maxboro/uncalibrated-visual-odometry-cpp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Description

Simplified visual odometry for uncalibrated camera path estimation in C++ using OpenCV.

Pipeline consists of:

  • Keypoint extraction using ORB
  • Matching keypoints between each pair of consecutive frames
  • Filtering matches using:
    • Descriptor distance ratio test (Lowe's ratio)
    • RANSAC inlier filtering based on epipolar geometry
  • Estimating the essential matrix from matched keypoints
  • Recovering relative camera pose (rotation and translation direction) from the essential matrix

How to run

Opencv installation

sudo apt install libopencv-dev

Set path to video

Path should be set via INPUT_VIDEO_PATH constexpr in main.cpp.

Build and run

make           # Builds the project
./main         # Runs the executable

Clean

make clean     # Cleans up build files

Example

Demo

Credits

OpenCV is used for most of image processing

About

Visual odometry on uncalibrated video

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published