This dataset is composed of photos obtained by attaching multiple ArUco markers to the walls in the corner of a darkened room, projecting a video containing moving shadows and lighting patterns over the corner, and then recording the resulting scene from 6 different points of view, for a total of 8652 frames with a resolution of 800 x 600. For each image, the ground-truth labels of the markers are provided, tagged with a method based on the classic ArUco technique, and manually adding the positions and IDs of undetected markers.

Examples

Some examples from the dataset are the following:

Dataset structure

The dataset is provided as six compressed folders, each representing a unique point of view. Every folder contains the individual frames obtained from that point of view, along with a subdirectory named «corrected_annotations.» For each frame, there is an accompanying JSON file with a matching name located in the «corrected_annotations» subdirectory containing the ground-truth positions (2D coordinates) of each marker in the image, as well as their ID. Please refer to the following example for clarification:

{
    "markers": [
        {
            "id": 238,
            "corners": [
                {
                    "x": 5.3789202880859375e+02,
                    "y": 2.0391471862792969e+02
                },
                {
                    "x": 5.8078918457031250e+02,
                    "y": 2.5667852783203125e+02
                },
                {
                    "x": 5.2534979248046875e+02,
                    "y": 2.8923260498046875e+02
                },
                {
                    "x": 4.8175805664062500e+02,
                    "y": 2.3331109619140625e+02
                }
            ]
        },
        ...

Download our dataset

Our dataset can be downloaded through Zenodo

Read our work

A preprint of our paper can be accessed through ArXiv: link

Citing

If you use this work in your research, you must cite:

1. Rafael Berral-Soler, Rafael Muñoz-Salinas, Rafael Medina-Carnicer, Manuel J. Marín-Jiménez, DeepArUco++: Improved detection of square fiducial markers in challenging lighting conditions, Image and Vision Computing, Volume 152, 2024, 105313, ISSN 0262-8856, https://doi.org/10.1016/j.imavis.2024.105313
2. Rafael Berral-Soler, Rafael Muñoz-Salinas, Rafael Medina-Carnicer, and Manuel J. Marín-Jiménez. 2023. DeepArUco: Marker Detection and Classification in Challenging Lighting Conditions. In: IbPRIA 2023. Lecture Notes in Computer Science, vol 14062. Springer, Cham. https://doi.org/10.1007/978-3-031-36616-1_16

Contact