An object/segment oriented skew-correction technique for document images

An innovative skew correction technique is presented. The method is not pixel oriented as most classical techniques, but is object/segment oriented. It does not presuppose a unique text orientation in the document image and can handle document images including "text" blocks with different orientations. It is very efficient in the sense that only the relevant data/information is rotated.

[1]  Rainer Hoch,et al.  Document analysis at DFKI. - Part 2: Information extraction , 1995 .

[2]  S.C. Hinds,et al.  A document skew detection method using run-length encoding and the Hough transform , 1990, [1990] Proceedings. 10th International Conference on Pattern Recognition.

[3]  Frank Hönes,et al.  Layout extraction of mixed mode documents , 2005, Machine Vision and Applications.

[4]  Henry S. Baird,et al.  The skew angle of printed documents , 1995 .

[5]  Yuan Yan Tang,et al.  Document skew detection based on the fractal and least squares method , 1995, Proceedings of 3rd International Conference on Document Analysis and Recognition.

[6]  Eberhard Mandler,et al.  One-pass encoding of connected components in multivalued images , 1990, [1990] Proceedings. 10th International Conference on Pattern Recognition.

[7]  Achim Weigel,et al.  Document analysis at DFKI. - Part 1: Image analysis and text recognition , 1995 .

[8]  Andreas Dengel,et al.  Rotationswinkelbestimmung in abgetasteten Dokumentbildern , 1989, DAGM-Symposium.