Directional Decomposition for Odia Character Recognition
Present work aims at analyzing the role of directional features for efficient recognition of printed Odia characters. The characteristics of Odia scripts that demand for separate rigorous OCR research are identified. Directional features are extracted by directional decomposition of character image and using fixed zoning. The zones are determined based on the input character patterns. Initial experiment with a modest size of 20 features are taken by considering 4 directions and 5 fixed zones yield very promising results. It is shown that these features yield nearly 95% recognition accuracy by multi-class SVM classifiers. High acuracy of recognition justifies the importance of directional decomposition which indirectly captures the stroke information of the script. In another experiment, we consider 164 dimensional feature vectors by taking zones for the entire image. The observation made here can prove to be useful in building an efficient OCR system for Odia characters.
KeywordsOCR of Indic Scripts Feature extraction Directional decomposition Odia scripts
Unable to display preview. Download preview PDF.
- 1.Bai, Z.-L., Huo, Q.: A study on the use of 8- Directional features for online handwritten character recognition. In: ICDAR 2005, pp. 262–266 (2005)Google Scholar
- 2.Chang, C.C., Lin, C.J.: LIBSVM: a library for support vector machines (2001), Software available at: http://www.csie.ntu.edu.tw/cjlin/libsvmS
- 3.Chaudhuri, B.B., Pal, U., Mitra, M.: Automatic recognition of printed Oriya script. In: SADHANA, Part 1, vol. 27, pp. 23–34 (2002)Google Scholar
- 4.Fujisawa, H., Liu, C.-L.: Directional Pattern Matching for Character Recognition Revisited. In: ICDAR, pp. 794–798. IEEE Computer Society (2003)Google Scholar
- 5.Gao, X., Jin, L.-W., Yin, J.-X., Huang, J.-C.: A New Stroke-Based Directional Feature Extraction Approach for Handwritten Chinese Character Recognition. In: ICDAR 2010, p. 635 (2001)Google Scholar
- 6.Harit, G., Chaudhury, S., Garg, R.: GFG-Based Compression and Retrieval of Document Images in Indian Scripts. In: Govindraju, Setlu (eds.) Guide to OCR for Indic Scripts, Advances in Pattern Recognition, pp. 269–284 (2010)Google Scholar
- 8.Negi, A., Bhagvati, C., Krishna, B.: An OCR system for Telugu. In: ICDAR 2001, pp. 1110–1114 (2001)Google Scholar
- 10.Pal, U., Sharma, N., Wakabayashi, T., Kimura, F.: OffLine Handwritten Character Recognition of Devanagari Script. In: Proc. 9th International Conference on Document Analysis and Recognition, pp. 496–500 (2007)Google Scholar