Offline Recognition of Malayalam and Kannada Handwritten Documents Using Deep Learning
For a variety of reasons, handwritten text can be digitalized. It is used in a variety of government entities, including banks, post offices, and archaeological departments. Handwriting recognition, on the other hand, is a difficult task as everyone has a different writing style. There are essentially two methods for handwritten recognition: a holistic and an analytic approach. The previous methods of handwriting recognition are time- consuming. However, as deep neural networks have progressed, the approach has become more straightforward than previous methods. Furthermore, the bulk of existing solutions are limited to a single language. To recognise multilanguage handwritten manuscripts offline, this work employs an analytic approach. It describes how to convert Malayalam and Kannada handwritten manuscripts into editable text. Lines are separated from the input document first. After that, word segmentation is performed. Finally, each word is broken down into individual characters. An artificial neural network is utilised for feature extraction and classification. After that, the result is converted to a word document.
Bhowmik T.K., Parui S.K., Roy U., (2008). Discrim- inative hmm training with ga for handwritten word recognition, 2008 19th International Conference on Pattern Recognition, IEEE, USA. https://doi.org/10.1109/ICPR.2008.4761830
Rahiman M.A., & Rajasree M., (2011). Recognition of simple and conjunct handwritten malayalam characters using lcpa algorithm, International Conference on Advances in Computing and Communications, Springer, Berlin. https://doi.org/10.1007/978-3-642-22720-2_31
Hashrin C., Jossy A., Sudhakaran K., Thushara A., & John A., (2019). Segmenting characters from malayalam hand- written documents, in 2019 1st International Confer- ence on Innovations in Information and Communication Technology (ICIICT), IEEE, India.
John J., Pramod K.V., & Balakrishnan K., (2012). Uncon- strained handwritten malayalam character recognition using wavelet transform and support vector machine classifier, Procedia Engineering, 30, 598–605. https://doi.org/10.1016/j.proeng.2012.01.904
Dhaka V.P., Sharma M.K., (2015). An efficient segmenta- tion technique for devanagari offline handwritten scripts using the feedforward neural network, Neural Computing and Applications, 26(8), 1881–1893. https://doi.org/10.1007/s00521-015-1844-9
Malakar S., Sharma P., Singh P.K., Das M., Sarkar R., & Nasipuri M., (2017). A holistic approach for handwritten hindi word recognition, International Journal of Com- puter Vision and Image Processing (IJCVIP), 7(1), 59–78. https://doi.org/10.4018/IJCVIP.2017010104
Saha S., Som T., (2011). Hand written character recognition using fuzzy membership function, IJETSE International Journal of Emerging Technologies in Sciences and Engineering, 5(2), 11-15. https://dx.doi.org/10.2139/ssrn.2009131.
Sahoo S., Nandi S.K., Barua S., Bhowmik S., Malakar S., Sarkar R., (2018). Handwritten bangla word recognition using negative refraction-based shape trans- formation, Journal of Intelligent & Fuzzy Systems, 35(2), 1765–1777. https://doi.org/10.3233/JIFS-169712
Cherkauer B.S., & Friedman E.G., (1995). A unified design methodology for cmos tapered buffers, IEEE Transactions on Very Large-Scale Integration (VLSI) Systems, 3(1), 99–111. https://doi.org/10.1109/92.365457
Das D., Nayak D. R., Dash R., Majhi B., & Y.-D. Zhang, (2020). H-wordnet: A holistic convolutional neural network approach for handwritten word recognition, IET Image Processing, 49(9), 1794-1805. https://doi.org/10.1049/iet-ipr.2019.1398
Bhowmik S., Malakar S., Sarkar R., & Nasipuri M., (2014). Handwritten bangla word recognition using elliptical features, International Conference on Com- putational Intelligence and Communication Networks, IEEE, India. 257–261. https://doi.org/10.1109/CICN.2014.66
Dasgupta J., Bhattacharya K., & Chanda B., (2016). A holistic approach for off-line handwritten cursive word recognition using directional feature based on arnold transform, Pattern Recognition Letters, 79(1) 73-79. https://doi.org/10.1049/iet-ipr.2019.1398
Shanjana, C., James, A., (2015). Offline recognition of malayalam handwritten text, Procedia Technology, 19,772–779. https://doi.org/10.1016/j.protcy.2015.02.109
Alex, Meenu & Das, Smija, (2016). An approach towards malayalam handwriting recognition using dissimilar classifiers, Procedia Technology, 25, 224–231. https://doi.org/10.1016/j.protcy.2016.08.101
Bag S., & Krishna A., (2015). Character segmentation of hindi unconstrained handwritten words, International workshop on combinatorial image analysis, Springer. 247–260. https://doi.org/10.1007/978-3-319-26145-4_18
Barua S., Malakar S., Bhowmik S., Sarkar R., Nasipuri M., (2017). Bangla hand written city name recognition using gradient-based feature, Proceedings of the 5th International Conference on Frontiers in Intelligent Computing: Theory and Applications, Springer, 515, 343–352. https://doi.org/10.1007/978-981-10-3153-3_34
Bhowmik S., Malakar S., Sarkar R., Basu S., Kundu M., & Nasipuri M., (2019). Off-line bangla handwritten word recognition: A holistic approach, Neural Computing and Applications, 31(10), 5783–5798. https://doi.org/10.1007/s00521-018-3389-1
Pramanik R., Bag S., (2020). Segmentation-based recog- nition system for handwritten bangla and devana- gari words using conventional classification and trans- fer learning, IET Image Processing, 14(5), 959–972. https://doi.org/10.4018/IJCVIP.2017010104
Tamen Z., Drias H., Boughaci D., An efficient multiple classifier system for arabic handwritten words recognition, Pattern Recognition Letters, 93(1), 123-132. https://doi.org/10.1016/j.patrec.2017.01.020
Copyright (c) 2021 Ayna Asokan, Sreeleja N Unnithan
This work is licensed under a Creative Commons Attribution 4.0 International License.
Views: Abstract : 60 | PDF : 10