chjxu / deeptag-pytorch Goto Github PK

Implementation of paper DeepTag: A General Framework for Fiducial Marker Design and Detection.

Project page: https://herohuyongtao.github.io/research/publications/deep-tag/.

DeepTag is a general framework for fiducial marker design and detection, which supports existing and newly-designed marker families. DeepTag is a two-stage marker detection pipeline:

• Stage-1: detect ROIs of potential markers;
• Stage-2: detect keypoints and digital symbols inside each ROI, then determine 6-DoF pose and marker ID.

• For image input:

  python test_deeptag.py --config config_image.json
  

• For video input:

  python test_deeptag.py --config config_video.json
  

The configuration file is in JSON format. Please modify the configurations to fit your needs. Example configurations files for image and video input are provided (i.e., config_image.json and config_video.json).

Detail explaination of configuration file:

• is_video: {0, 1} for image/video respectively.
• filepath: path of input image/video (use 0 for webcam input).
• family: marker family, currently support {apriltag, aruco, artoolkitplus, runetag, topotag, apriltagxo}.
• hamming_dist: Hamming dist for checking the marker library; normally, 4 works well enough.
• codebook: path of codebook; if it is empty, the default path codebook/FAMILY_codebook.txt will be used. For markers with multiple codebooks like AprilTag and ArUco, their default codebooks are for AprilTag (36h11) and ArUco (36h12) respectively.
• cameraMatrix: camera intrinsic matrix, [fx, 0, cx, 0, fy, cy, 0, 0, 1].
• distCoeffs: camera distortion coefficients (both radial and tangential), [k1, k2, p1, p2, k3, k4, k5, k6].
• marker_size: physical size of the marker.

Besides supporting existing markers like AprilTag, ArUco, ARToolkitPlus, TopoTag & RuneTag, DeepTag also supports newly-designed markers like AprilTag-XO, AprilTag-XA and RuneTag+ (provided in folders images_tag). Set family to apriltagxo in config for AprilTag-XO and AprilTag-XA, and runetag for RuneTag+ respectively.

The source code is provided for research purposes only. Any commercial use is prohibited. When using the code in your research work, please cite the following paper:

"DeepTag: A General Framework for Fiducial Marker Design and Detection."
Zhuming Zhang, Yongtao Hu, Guoxing Yu, and Jingwen Dai
IEEE TPAMI 2023.

@article{zhang2023deeptag,
  title={{DeepTag: A General Framework for Fiducial Marker Design and Detection}},
  author={Zhang, Zhuming and Hu, Yongtao and Yu, Guoxing and Dai, Jingwen},
  journal={IEEE Transactions on Pattern Analysis and Machine Intelligence},
  volume={45},
  number={3},
  pages={2931-2944},
  year={2023},
  publisher={IEEE}
}

If you find any bug or have any question about the code, please report to the Issues page.