Multiple classifier for degraded machine printed character recognition

The general problem of optical character recognition (OCR) remains a fundamental but not entirely solved issue in document analysis. In spite of significant improvements in the area of optical character recognition, the recognition of degraded printed characters, in particular, is still lacking satisfactory solutions. This paper presents an OCR method that combines the Hopfield network with a set of autoassociators for degraded character recognition. In the serial combination, the first classifier must achieve lower errors and be very well suited for rejection, whereas the second classifier must allow only low errors and rejects. A relative distance is used as a quality measurement parameter which makes the Hopfieldbased classifier very powerful and very well suited for rejection. We report experimental results for a comparison of three methods: the Hopfield model, the autoassociator-based classifier and the proposed combined architecture. Experimental results show the ability of the model to yield relevant and robust recognition with no errors on poor quality bank check characters even when the patterns are highly degraded. Mots-clés : Hopfield model, associative memory, degraded printed characters, autoassociator, OCR, serial combination, character recognition..

[1]  Berrin A. Yanikoglu,et al.  Pitch-based segmentation and recognition of dot-matrix text , 2000, International Journal on Document Analysis and Recognition.

[2]  Abderrezak Guessoum,et al.  Hopfield-multilayer-perceptron serial combination for accurate degraded printed character recognition , 2006 .

[3]  Giovanni Soda,et al.  A serial combination of connectionist-based classifiers for OCR , 2001, International Journal on Document Analysis and Recognition.

[4]  Zheru Chi,et al.  Handwritten numeral recognition using self-organizing maps and fuzzy rules , 1995, Pattern Recognit..

[5]  Chew Lim Tan,et al.  Imaged Document Text Retrieval Without OCR , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  B. Kapralos,et al.  An introduction to digital image processing , 1990 .

[7]  Yingbai Yan,et al.  Postprocessing algorithm for the optical recognition of degraded characters , 1999, Electronic Imaging.

[8]  Madjid Arezki,et al.  Sequential neural network combination for degraded machine-printed character recognition , 2005, IS&T/SPIE Electronic Imaging.

[9]  George Nagy,et al.  Twenty Years of Document Image Analysis in PAMI , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  Sebastiano Impedovo,et al.  Automatic Bankcheck Processing: A New Engineered System , 1997, Int. J. Pattern Recognit. Artif. Intell..

[11]  Tin Kam Ho,et al.  Enhancing degraded document images via bitmap clustering and averaging , 1997, Proceedings of the Fourth International Conference on Document Analysis and Recognition.

[12]  F T Yu,et al.  Two-dimensional programmable optical neural network. , 1989, Applied optics.

[13]  Tianxu Zhang,et al.  Rotation and scale change invariant point pattern relaxation matching by the Hopfield neural network , 1997 .

[14]  Norihiro Hagita,et al.  Text-Line Extraction and Character Recognition of Document Headlines With Graphical Designs Using Complementary Similarity Measure , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  S. Hyakin,et al.  Neural Networks: A Comprehensive Foundation , 1994 .