Published June 4, 2019 | Version 1.0
Dataset Open

Photometric stereo data of non-crimp fabric boundaries




This data set was acquired in the context of EU project ZAero. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 721362. Project duration: 2016/10/01 - 2019/09/30. This data set contains a set of 68 HDF5 files from two different example surface patches of NCF carbon fiber fabrics.

For more information about the HDF5 format, please visit the HDF5 Group website:

Each file contains data from a photometric stereo vision system [1]. The data set is intended for evaluation of methods for carbon fiber fabric edge detection [2].

Each file contains a single 3-dimensional (W x H x L) array 'raw' with:
H...height of the images
W...width of the images
L...number of images with different light sources (L = 8)

An example for loading and visualizing the data in Python comes with this data set:


  author    = {Palfinger, Werner and Thumfart, Stefan and Eitzinger, Christian},
  title     = {Photometric stereo on carbon fiber surfaces},
  booktitle = {35th Workshop of the Austrian Association for Pattern Recognition},
  year      = {2011}

  author    = {Zambal, Sebastian and Heindl, Christoph and Eitzinger, Christian}
  title     = {Probabilistic Modelling combined with a CNN for boundary detection of carbon fiber fabrics},
  booktitle = {IEEE International Conference on Industrial Informatics (INDIN)}
  year      = {2019}




Files (610.1 MB)

Name Size Download all
610.1 MB Preview Download

Additional details


ZAero – Zero-defect manufacturing of composite parts in the aerospace industry 721362
European Commission