The TrigraphSlant (Img) data set, March 2011

This data set contains images of handwriting, produced under conditions of normal and disguised slant.

It can be used to evaluate writer-verification algorithms or the recognition performance of a reading system.
The example image below shows two samples of handwriting from a single writer in natural slant (top row). The bottom row shows two samples which are disguised with a leftward (bottom left) and a rightward slant (bottom right), respectively. Results on this data set are reported in article: A.A. Brink, R.M.J. Niels, R.A. van Batenburg, C.E. van den Heuvel, and L.R.B. Schomaker, Towards robust writer verification by correcting unnatural slant, Pattern Recognition Letters, Volume 32, Issue 3, 1 February 2011, Pages 449-457, ISSN 0167-8655, DOI: 10.1016/j.patrec.2010.10.010.

Download (right mouse button)
Please read the README for conditions of use.