Gustaf Kylberg

back to main page

Kylberg Sintorn Rotation dataset

Figure 1. Example patches from each one of the 25 texture classes.

Short description

  • 25 texture classes, see Figure 1.
  • 100 samples per texture class.
  • Samples are 122x122 pixels in size.
  • The textures has been rotated using hardware and rotated by interpolation using, nearest neighbour, linear, 3rd order cubic, B-spline, and Lanczos 3 kernels.
  • All texture samples are normalized with a mean value of 127 and a standard deviation of 40.


Kylberg Sintorn Rotation dataset

The compressed files are roughly 240 MB in size each.

  • Hardware rotated texture samples. [.zip, .7z]
  • Rotated texture samples using nearest neighbour. [.zip, .7z]
  • Rotated texture samples using linear interpolation. [.zip, .7z]
  • Rotated texture samples using 3rd order cubic interpolation. [.zip, .7z]
  • Rotated texture samples using B-spline interpolation. [.zip, .7z]
  • Rotated texture samples using Lanczos 3 interpolation. [.zip, .7z]

Original Images


  1. Kylberg G. Automatic Virus Identification using TEM - Image Segmentation and Texture Analysis
    PhD thesis at Uppsala University, March 2014.
    [abstract and full text]
  2. Gonzalez, E.; Fernandez, A. & Bianconi, F. General Framework for Rotation Invariant Texture Classification Through Co-occurrence of Patterns
    Journal of Mathematical Imaging and Vision, 2014, 1-14.
    [abstract] [doi]
  3. Bianconi, F. & Fernandez, A. An appendix to "Texture databases - A comprehensive survey"
    Pattern Recognition Letters, 2014, 1-14.
    [abstract] [doi]


A photo from the setup during acquisition.