ArUCo markers and Computer Vision for Localisation

This module was modified from ArUCo-Markers-Pose-Estimation-Generation-Python. The original license still applies to this code regardless of these modifications.

https://github.com/user-attachments/assets/f96c039b-bd86-4312-868f-6efccbbd9363

Purpose

The purpose of this module is to convert a video into an array of locations & timestamps like this:

[{"location": [108.404, -0.314, 172.88], "timestamp": "2025-09-13 17:59:31.348000"}, 
{"location": [108.378, -0.304, 172.877], "timestamp": "2025-09-13 17:59:31.414000"}, 
...
{"location": [109.216, -0.114, 173.852], "timestamp": "2025-09-13 18:00:02.558000"}]

Equipment Required

• a camera high-quality camera = more precision further away from camera
• a printed ArUCo tag
• a chessboard for calibration printed on paper or a real game board

Setup Repository

Create a virtual environment Windows: python -m venv venv Activate virtual environment Windows: venv/Scripts/activate Install dependencies pip install -r requirements.txt Setup Done, start localising below 🙃

Localisation Method

Step 0: Calibrate camera

This means to find out your current camera calibration including distortion (skew, round distortion) and intrinsic camera properties (focal length, principal point). The calibration output is two matrices which are saved as intrinsic_matrix.npy and distortion_coefficients.npy.

To calibrate your camera, do: a. take a video of a chessboard b. save the video under previously_calibrated_cameras c. open calibrate_camera.py d. adjust camera_filename, chessboard_size to match e. run python calibrate_camera.py

Step 1: Place your camera

During localisation the camera needs to be fixed in the same place. It is ok if the target is moving. Notice where you place your camera i.e. what is the translation and rotation relative to a origin (e.g. corner of building). Adjust these variables in the measure_locations.py file:

camera_translation = [103.849, 0, 183.396]   # in metres from origin
camera_rotation = [0, 90, 0]
camera = Camera(load_calibration_details_for_camera_name("My Smartphone Camera"),
                camera_translation, camera_rotation)

Step 2: State your ArUCo tag

State how which type of ArUCo tag you printed out and how big it is in cm:

aruco_dict_type = "DICT_5X5_100"
aruco_marker_length = 15.9

If you don't have a ArUCo tag, use the following command to generate one and then print it out after:

python ./aruco_markers_computer_vision/generate_aruco_tags.py --id 24 --type DICT_5X5_100 --output tags/

Step 3: Record localisation video

Placing your calibrated camera in its correct spot, record a video and specify its location and when you started recording:

video_start_time = datetime(2025, 9, 18, 14, 55, 28)
video_file_path = f"./calculating_location/original_source_videos/{video_name}"

Step 4: Run python measure_locations.py

By default, the locations are output to calculating_location/output folder.

Notes

If you found this repository useful, take a look at and star ⭐️ the original repository.

Feel free to reach out to me or the original author:

Me - arnard76@gmail.com or ashe292@aucklanduni.ac.nz

Original Author - github.com/GSNCodes

Happy Learning! Keep chasing your dreams!

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

Common Issues

OpenCV parse "warning"

Shows during pose estimation / measuring locations, but doesn't stop the actual thing from running so just wait patiently.

indoor-localisation-accuracy-testing\ArUCo-markers-cv-localisation-method\calculating_location\pose_estimation.py:96: RuntimeWarning: invalid value encountered in cast
  imgpts = imgpts.reshape(-1, 2).astype(int)
OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type

OpenCV(4.12.0) :-1: error: (-5:Bad argument) in function 'line'
> Overload resolution failed:
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type
>  - Can't parse 'pt2'. Sequence item with index 0 has a wrong type