Depart. of Information &
Communication Systems Engineering.
University of the Aegean Lymperis Building: Office 17 Karlovasi, 83200, Samos, Greece.

+30 22730 82263
+30 22730 82009
kavallieratou at

Today is: 24th September 2017

Evaluation Data For Binarization Algorithms

The ideal way of evaluation should be able to decide, for each pixel, if it has finally succeeded the right color (black or white) after the binarization. This is an easy task for a human observer but very difficult for a computer to perform it automatically for all the pixels of several images.

Two sets of 150 document images each were built up by using image mosaicing superimposing techniques for blending. In more detail, we used as target images pdfs and resized all the noisy images to A4 size. Then, we used two different techniques for the blending: the maximum intensity and the image averaging approaches.

At the maximum intensity technique (max_int), the new image was constructed by picking up for each pixel in the new image, the darkest corresponding pixel of the two images.

These synethetic images are here.
At the image averaging technique (ave_int), each pixel in the new image is the average of the two corresponding ones in the original images. These synethetic images are here.

You can find more information about the noise in the images, the techniques and existed results in the paper:

Pavlos Stathis, Ergina Kavallieratou, Nikos Papamarkos, "An Evaluation Technique for Binarization Algorithms", Journal of Universal Computer Science, Vol. 14, No. 18, pp. 3011-3030, 2009.

The database has also been used in the contests:
ICFHR 2010 Contest: Quantitative Evaluation of Binarization Algorithms
ICDAR 2011: Specific Document Analysis Algorithm Contributions in End-to-End Applications
For any remarks please contact: kavallieratou at

Copyright ©2013-2014 Elina Tsakalidou