THE IMPROVEMENT OF THE ALGORITHM OF RECOGNITION OF HANDWRITTEN TEXTS ON THE BASIS OF THE VALUATION IMAGE

One of the most popular approaches for handwriting recognition is the representation of entire words in sequences of symbols of the Markov chain. The set extracted from the images of the symbols is analyzed for compliance with a pre-prepared word patterns (model templates). The word whose model has the highest probability of formation of the analyzed sequences recognized the target. The variability of cursive writing words leads to the need of the analysis extracted from the image sequence of characters with models generated for words consisting of different numbers of digits. In the case where the analyzed word different from the word used for the model-only template, model template longer words earns you a mathematical advantage over the model of the shorter words, leading to recognition errors. The paper proposes to reduce recognition errors, the normalization of the image.

1. Horst Bunke Recognition of Cursive Roman Handwriting - Past, Present and Future // Proceedings of the Seventh International Conference on Document Analysis and Recognition (ICDAR 2003), 2003, Volume 1, pp. 448–459.

2. Norris, D. Shortlist B: A Bayesian Model of Continuous Speech Recognition // Psychological Review, Vol. 115, No. 2, 2008 pp. 357–395.

3. Mozgovoy A.A. Problemy primeneniya skrytykh markovskikh modeley pri raspoznavanii rukopisnogo teksta // V mire nauchnykh otkrytiy. 2013. No.6. pp.186-198.

4. S. Sangeetha Devi, Dr. T. Amitha Invariant and Zernike Based Offline Handwritten Character Recognition // International Journal of Advanced Research in Computer Engineering & Technology (IJARCET) Volume 3 Issue 5, May 2014, pp. 1950-1954.

5. Mozgovoy, A.A. Metodika sinteza slovarya dlya zadachi avtomaticheskogo raspoznavaniya rukopisnykh slov // Telekommunikatsii. 2014. No.5. pp.3-4.

6. Louloudis G., Gatos B., Halatsis C. Text Line Detection in Unconstrained Handwritten Documents Using a BlockBased Hough Transform Approach // Document Analysis and Recognition, 2007. ICDAR 2007. Ninth International Conference on, Volume 2. pp. 599-603.

7. Vijay Laxmi Sahu, Babita Kubde Offline Handwritten Character Recognition Techniques using Neural Network: A Review IJSR Volume 2 Issue 1, January 2013 pp. 87-94.