Integrating Writing Direction and Handwriting Letter Recognition in Touch-Enabled Devices
Optical character recognition (OCR) transforms printed text to editable format and digital writing on smart devices. Learning to write programs has made learners trace an alphabet to learn the flow of writing and OCR by itself is less effective as it ignores the directional flow of writing and only focuses on the final image. Our research designed a unique android-based multilingual game-like writing app that enhances the writing experience. A key focus of the research was to compare and identify character recognition algorithms that are effective on low-cost android tablets with limited processing capabilities. We integrate a quadrant-based direction checking system with artificial neural networks and compare it to the existing systems. Our solution has the dual advantage of evaluating the writing direction and significantly increasing the accuracy compared to the existing systems. This program is used as the literacy tool in many villages in rural India.
KeywordsCharacter recognition Online recognition Offline recognition Neural networks Quadrant-based direction checking Alphabet rules Self-organizing map ANN OCR
This work derives its inspiration and direction from the Chancellor of Amrita University, Sri Mata Amritanandamayi Devi. We are grateful for the support of our colleagues at Amrita CREATE and the staff at Amrita University.
- 2.Nedungadi, P., Jayakumar, A., Raman, R.: Low cost tablet enhanced pedagogy for early grade reading: Indian context. In: Humanitarian Technology Conference (R10-HTC), 2014 IEEE Region 10, pp. 35–39. IEEE, Aug 2014Google Scholar
- 4.Cha, S.H., Srihari, S.N.: Writing speed and writing sequence invariant on-line handwriting recognition. Pattern Recogn. 10, 9789812386533_0020 (2001)Google Scholar
- 7.Shaw, B., Kumar Parui, S., Shridhar, M.: Offline handwritten devanagari word recognition: A holistic approach based on directional chain code feature and HMM. In: International Conference on Information Technology, ICIT’08, pp. 203–208. IEEE (2008)Google Scholar
- 8.Pal, U., Wakabayashi, T., Kimura, F.: Comparative study of Devnagari handwritten character recognition using different feature and classifiers. In: 10th International Conference on Document Analysis and Recognition, ICDAR’09, pp. 1111–1115. IEEE (2009)Google Scholar
- 9.Arora, S., Bhattacharjee, D., Nasipuri, M., Basu, D.K., Kundu, M.: Recognition of non-compound handwritten devnagari characters using a combination of mlp and minimum edit distance. arXiv preprint arXiv:1006.5908 (2010)Google Scholar
- 10.Kompalli, S., Nayak, S., Setlur, S., Govindaraju, V. Challenges in OCR of Dev anagari Documents. In: ICDAR, pp. 327–333. Aug (2005)Google Scholar
- 11.Heaton, J.: Introduction to Neural Networks with Java. Heaton Research, Inc (2008)Google Scholar
- 12.Sharma, K.S., Karwankar, A.R., Bhalchandra, A.S.: Devnagari character recognition using self organizing maps. In: 2010 IEEE International Conference on Communication Control and Computing Technologies (ICCCCT). pp. 687–691. IEEE Oct 2010Google Scholar
- 13.Araokar, S.: Visual character recognition using artificial neural networks.arXiv preprint cs/050501 (2005)Google Scholar