ArUco_Pose_Estimation

This repository contains the bare minimum code for simple pose estimation using ArUco markers in Python

Dependencies

• OpenCV
• Numpy

Installation

1. Download the repository:

   git clone https://github.com/Jshulgach/ArUco_Pose_Estimation.git
   cd ArUco_Pose_Estimation

2. Install dependencies according to your OS:

   Windows

   It's best to install dependencies in a virtual environment. Using either Anaconda or Python venv, prepare your environment:

   python -m venv aruco python=3.10
   call aruco\Scripts\activate
   pip install -r requirements.txt

   Linux

   Install using the apt install command:

   sudo apt install python3-opencv python3-yaml

Usage

Checkerboard

Make sure to have a checkerboard printed if you want to have accurate pose calibration. You can find a multitude of patterns from Mark Hedley Jones. Once you print out the one you like, make sure to update the calibration_settings.yaml file with the correct dimensions of the squares on the checkerboard.

Running the scripts

The files are named in the order of operations:

• 1_save_frames.py : Used to save frames from the camera feed.
• 2_calibrate.py : Performs the calibration routine with a single camera, computes distortion matrix.
• 3_main.py : Detects ArUco markers in the camera feed.

You can find more details on parameters for each script using python my-script-to-run.py --help. Ideally all of these should be run without much to change

1. Save Calibration Frames

   Run 1_save_frames.py to initialize the camera feed. Press the space bar when prompted to start collecting images of your checkerboard. Make sure to move the checkerboard around to get different poses and orientations.

   python 1_save_frames.py

2. Calibration
   Run 2_calibrate.py to read the checkerboard images in your directory and generate a calibration_matrix.npy and distortion_coefficients.npy file.

   python 2_calibrate.py  

   • Note: If you're connected to another device via ssh (like a raspberry pi) and need to transfer the .npy files, you can copy them to your local WSL environment with the command:

     scp myusername@192.168.1.164:/home/myusername/github/ArUco_Pose_Estimation/distortion_coefficients.npy /home/myusername

     For copying to your host windows maching:

     scp myusername@192.168.1.164:/home/myusername/github/ArUco_Pose_Estimation/distortion_coefficients.npy /mnt/c/Users/WindowsUserName

3. Pose Estimation
   Run 3_main.py to begin running the pose estimation for each ArUCo marker detected. This is done in real-time for each frame obtained from the webcam feed.

   python 3_main.py  

Feel free to reach out to me in case of any issues.
If you find this repo useful in any way please do star ⭐️ it so that others can reap it's benefits as well.

Acknowledgements

This repository is inspired by the work of GSNCodes

References

1. https://docs.opencv.org/4.x/d9/d6d/tutorial_table_of_content_aruco.html
2. https://docs.opencv.org/4.x/dc/dbb/tutorial_py_calibration.html