Online Farsi Handwritten Character Recognition Using Hidden Markov Model

In this paper, a method for recognizing online Farsi characters that are written separately has been introduced. Regarding to the shape and the structure of the main body, Farsi letters have been divided into 18 groups. First, hidden Markov model (HMM) technique has been exploited to recognize the main body. In the next step, the final recognition in each group is performed according to delayed strokes (dots and small signs) and their hidden Markov models. The proposed method has been tested on TMU dataset and the recognition accuracy of 95.9% and 94.2% has been obtained for the recognition of the group and the character, respectively.

[1]  Ehsanollah Kabir,et al.  Fusion of HMM Classifiers, Based on x, y and (x, y) Signals, for the Recognition of Online Farsi Handwriting: a Large Lexicon Approach , 2014 .

[2]  Marcus Liwicki,et al.  Feature Selection for HMM and BLSTM Based Handwriting Recognition of Whiteboard Notes , 2009, Int. J. Pattern Recognit. Artif. Intell..

[3]  Vahid Ghods,et al.  Online Farsi digit recognition using their upper half structure , 2015, International Conference on Graphic and Image Processing.

[4]  Saeed Bagheri Shouraki,et al.  Recognition of Persian Online Handwriting Using Elastic Fuzzy Pattern Recognition , 2007, Int. J. Pattern Recognit. Artif. Intell..

[5]  Sargur N. Srihari,et al.  On-Line and Off-Line Handwriting Recognition: A Comprehensive Survey , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Ehsanollah Kabir,et al.  A STUDY ON FARSI HANDWRITING STYLES FOR ONLINE RECOGNITION , 2013 .

[7]  Ehsanollah Kabir,et al.  Decision fusion of horizontal and vertical trajectories for recognition of online Farsi subwords , 2013, Eng. Appl. Artif. Intell..

[8]  Adel M. Alimi,et al.  Online Arabic handwriting recognition: a survey , 2013, International Journal on Document Analysis and Recognition (IJDAR).

[9]  Ehsanollah Kabir,et al.  Effect of delayed strokes on the recognition of online Farsi handwriting , 2013, Pattern Recognit. Lett..

[10]  Ehsanollah Kabir,et al.  Feature Extraction for Online Farsi Characters , 2010, 2010 12th International Conference on Frontiers in Handwriting Recognition.

[11]  Ehsanollah Kabir,et al.  Lexicon Reduction Using Delayed Strokes for the Recognition of Online Farsi Subwords , 2011 .