A Novel Skew Detection and Correction Approach for Scanned Documents

Riaz Ahmad; Muhammad Zeshan Afzal; Sheikh Faisal Rashid; Marcus Liwicki; Thomas Breuel; Andreas Dengel

In: DAS. IAPR International Workshop on Document Analysis Systems (DAS-12), April 11-14, Santorini, Greece, DAS, 4/2016.


Document analysis is relying on pre-processing stage. As much as we get optimal results in pre-processing, better will be the chances of analyzing the contents of a document. Similarly, for scanned text documents, a skew detection and then correction is very important, because, it significantly improves the accuracies in Optical Character Recognition (OCR) systems. This paper presents a novel approach for skew detection and correction. The approach is specifically tested on Latin and Arabic like scripts and outperform state of the art approaches.

Weitere Links

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz