Authors : Sasmita Kanhar; Shradhanjali Digal; Poojitha Das; Snehashree Pradhan; Priyanka Munda

Volume/Issue : Volume 11 - 2026, Issue 5 - May

Google Scholar : https://tinyurl.com/wk38nmsp

Scribd : https://tinyurl.com/mu63uymr

DOI : https://doi.org/10.38124/ijisrt/26May1173

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : Offline handwritten character recognition has gained significant attention in the fields of artificial intelligence and pattern recognition because of its applications in intelligent document analysis, banking automation, postal systems, educational platforms, and biometric authentication. The recognition of handwritten characters remains a challenging task due to variations in writing styles, stroke orientation, character deformation, background noise, and image distortion. Traditional Artificial Neural Network (ANN)-based approaches often depend on handcrafted feature extraction methods and exhibit limited robustness when applied to large and diverse handwritten datasets. To address these limitations, this paper proposes an Attention-Enhanced Convolutional Neural Network (AE-CNN) framework for offline handwritten character recognition. The proposed system integrates adaptive image preprocessing, data augmentation, convolutional feature extraction, and attention-guided learning to improve classification accuracy and feature discrimination. Initially, a baseline ANN model is implemented to evaluate the limitations of conventional neural architectures. Subsequently, a deep CNN model integrated with an attention mechanism is developed to improve recognition capability for visually similar handwritten characters. Experimental evaluation is performed using the EMNIST handwritten dataset. The obtained results demonstrate that the proposed AE-CNN framework achieves improved classification performance, enhanced generalization capability, and better robustness against noisy handwritten samples compared to traditional ANN and conventional CNN models. The proposed approach provides an efficient and scalable solution for intelligent handwritten recognition systems used in real-world document processing applications.

Keywords : Handwritten Character Recognition, Deep Learning, Artificial Neural Network, Convolutional Neural Network, Attention Mechanism, EMNIST Dataset, Pattern Recognition.

References :

1. R. C. Gonzalez and R. E. Woods, Digital Image Processing, 3rd ed. Pearson Education, 2008.
2. C. M. Bishop, Pattern Recognition and Machine Learning. Springer, 2006.
3. A. K. Jain, R. P. W. Duin, and J. Mao, “Statistical pattern recognition: A review,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 1, pp. 4–37, 2000.
4. S. Haykin, Neural Networks and Learning Machines, 3rd ed. Pearson, 2009.
5. Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE, vol. 86, no. 11, pp. 2278–2324, 1998.
6. G. Cohen, S. Afshar, J. Tapson, and A. van Schaik, “EMNIST: Extending MNIST to handwritten letters,” in International Joint Conference on Neural Networks, 2017.
7. A. Vaswani et al., “Attention is all you need,” in Advances in Neural Information Processing Systems, 2017.
8. A. K. Jain, R. P. W. Duin, and J. Mao, “Statistical pattern recognition: A review,” IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000.
9. L. R. Rabiner,A tutorial on hidden Markov models and selected applications in speech recognition,” Proceedings of the IEEE, 1989.
10. P. Dutta, J. S. Banerjee, and S. Bhattacharyya, “Handwritten character recognition using neural networks,” International Journal of Computer Applications, 2012.
11. A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” NIPS, 2012.
12. G. Cohen, S. Afshar, J. Tapson, and A. van Schaik,“EMNIST: Extending MNIST to handwritten letters,” IEEE, 2017.
13. Z. Zhong, L. Jin, and Z. Xie, “High performance offline handwritten Chinese character recognition using deep convolutional neural networks,” ICDAR, 2015.