Enhancing Script Identification in Dravidian Languages using Ensemble of Deep and Texture Features
DOI:
https://doi.org/10.71426/jcdt.v1.i1.pp50-58Keywords:
Script languages, Data processing, Convolutional Neural Network (CNN), Histogram of Oriented Gradients (HOG), Support vector machine (SVM), Local binary patterns (LBP)Abstract
Dravidian languages, including Tamil, Telugu, Kannada, and Malayalam, have complex orthographic structures, making script identification challenging particularly for camera-based document images. This study proposes a hybrid approach that combines deep learning and texture-based methods for robust script recognition. The GoogLeNet convolutional neural network (CNN) model is used to extract deep features, while local binary patterns (LBP) and histogram of oriented gradients (HOG) capture texture characteristics. These features are fused and classified using support vector machine (SVM) classifier. Results show that CNN features alone achieve 84.50% accuracy, LBP achieves 85.90%, and HOG achieves 76.10%, while their fusion significantly improves accuracy to 92.10%. The combination of CNN and HOG features reaches 95.00% accuracy, demonstrating the effectiveness of integrating deep learning with texture-based approaches. This method has applications in OCR systems and assistive technologies for the visually impaired.
References
[1] Steever SB. Introduction to the Dravidian languages. In The Dravidian languages 2019 Dec 18 (pp. 1-44). Routledge.
[2] Konya IV, Mare B. Adaptive methods for robust document image understanding (Doctoral dissertation, Universitäts-und Landesbibliothek Bonn). 10.1109/TKDE.2022.3192842
[3] Yang K, Yi J, Chen A, Liu J, Chen W, Jin Z. ConvPatchTrans: A script identification network with global and local semantics deeply integrated. Engineering Applications of Artificial Intelligence. 2022 Aug 1;113:104916. 10.1016/j.engappai.2022.104916
[4] Ghalati MK, Nunes A, Ferreira H, Serranho P, Bernardes R. Texture analysis and its applications in biomedical imaging: A survey. IEEE Reviews in Biomedical Engineering. 2021 Sep 27;15:222-46. 10.1109/RBME.2021.3115703
[5] Rane J, Mallick SK, Kaya O, Rane NL. Scalable and adaptive deep learning algorithms for largescale machine learning systems. Future Research Opportunities for Artificial Intelligence in Industry. 2024;4:39-92.
[6] Roy SK, Dubey SR, Chanda B, Chaudhuri BB, Ghosh DK. Texfusionnet: an ensemble of deep cnn feature for texture classification. InProceedings of 3rd International Conference on Computer Vision and Image Processing: CVIP 2018, Volume 2 2020 (pp. 271-283). Springer Singapore. 10.1007/978-981-32-9291-8_22
[7] Mallappa S, Dhandra BV, Mukarambi G. Hybridization of texture features for identification of Bi-lingual scripts from camera images at wordlevel. InComputer Vision and Machine Intelligence Paradigms for SDGs: Select Proceedings of ICRTAC-CVMIP 2021 2023 Jan 1 (pp. 113-124). Singapore: Springer Nature Singapore.
10.1007/978-981-19-7169-3_11
[8] Gupta MK, Dhawan S, Kumar A. Document Image Script Identification using Deep Network. In2024 11th International Conference on Signal Processing and Integrated Networks (SPIN) 2024 Mar 21 (pp. 174-179). IEEE. 10.1109/SPIN60856.2024.10511557
[9] Naosekpam V, Sahu N. A hybrid scene text script identification network for regional Indian languages. ACM Transactions on Asian and Low-Resource Language Information Processing. 2024 Aug 8;23(8):1-26. https://doi.org/10.1145/36494
[10] Li X, Zhan H, Shivakumara P, Pal U, Lu Y. SANet-SI: A new self-attention-network for script identification in scene images. Pattern Recognition Letters. 2023 Jul 1;171:45-52. 10.1016/j.patrec.2023.04.015
[11] Nicolaou A, Bagdanov AD, Gomez L, Karatzas D. Visual script and language identification. In2016 12th IAPR workshop on document analysis systems (DAS) 2016 Apr 11 (pp. 393-398). IEEE. 10.1109/DAS.2016.63
[12] Li L, Tan CL. Script identification of camera-based images. In2008 19th International Conference on Pattern Recognition 2008 Dec 8 (pp. 1-4). IEEE. 10.1109/ICPR.2008.4760965
[13] Bhunia AK, Konwer A, Bhunia AK, Bhowmick A, Roy PP, Pal U. Script identification in natural scene image and video frames using an attention based convolutional-LSTM network. Pattern Recognition. 2019 Jan 1;85:172-84. https://doi.org/10.1016/j.patcog.2018.07.034
[14] Jajoo M, Chakraborty N, Mollah AF, Basu S, Sarkar R. Script identification from camera-captured multi-script scene text components. InRecent Developments in Machine Learning and Data Analytics: IC3 2018 2019 (pp. 159-166). Springer Singapore. 10.1007/978-981-13-1280-9_16
[15] Fasil OK, Manjunath S, Aradhya VM. Word-level script identification from scene images. In Proceedings of the 5th International Conference on Frontiers in Intelligent Computing: Theory and Applications: FICTA 2016, Volume 2 2017 (pp. 417-426). Springer Singapore. 10.1007/978-981-10-3156-4_43
[16] Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A. Going deeper with convolutions. In Proceedings of the IEEE conference on computer vision and pattern recognition 2015 (pp. 1-9). 10.1109/CVPR.2015.7298594.
[17] Li G, Muller M, Thabet A, Ghanem B. DeepGCNs: Can GCNs go as deep as CNNs?. In Proceedings of the IEEE/CVF international conference on computer vision 2019 (pp. 9267-9276). http://dx.doi.org/10.1109/ICCV.2019.00936.
[18] Touvron H, Cord M, Sablayrolles A, Synnaeve G, Jégou H. Going deeper with image transformers. In Proceedings of the IEEE/CVF international conference on computer vision 2021 (pp. 32-42). https://doi.org/10.48550/arXiv.2103.17239
[19] Kishore A, Singh S. Natural language image descriptor. In 2015 IEEE Recent Advances in Intelligent Computational Systems (RAICS) 2015 Dec 10 (pp. 110-115). IEEE. 10.1109/RAICS.2015.7488398
[20] Ojala T, Pietikainen M, Maenpaa T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns. IEEE Transactions on pattern analysis and machine intelligence. 2002 Aug 7;24(7):971-87. 10.1109/TPAMI.2002.1017623
[21] García-Olalla O, Alegre E, Fernández-Robles L, García-Ordás MT, García-Ordás D. Adaptive local binary pattern with oriented standard deviation (ALBPS) for texture classification. EURASIP journal on image and video processing. 2013 Dec;2013:1-1. 10.1186/1687-5281-2013-31
[22] Dalal N, Triggs B. Histograms of oriented gradients for human detection. In2005 IEEE computer society conference on computer vision and pattern recognition (CVPR'05) 2005 Jun 20 (Vol. 1, pp. 886-893). IEEE. 10.1109/CVPR.2005.177
[23] Déniz O, Bueno G, Salido J, De la Torre F. Face recognition using histograms of oriented gradients. Pattern recognition letters. 2011 Sep 1;32(12):1598-603. 10.1016/j.patrec.2011.01.004
[24] Rane N, Choudhary SP, Rane J. Ensemble deep learning and machine learning: applications, opportunities, challenges, and future directions. Studies in Medical and Health Sciences. 2024 Jul 4;1(2):18-41. 10.48185/smhs.v1i2.1225
[25] H. K. Bhargav, Ambresh Bhadrashetty, K. Neelashetty, V. B. Murali Krishna, G. Manohar Bali. A Graph-Based and Pattern Classification Approach for Kannada Handwritten Text Recognition Under Struck-Out Conditions. IJCESEN [Internet]. 2025 Mar. 4 [cited 2025 Jun. 30];11(1). Available from: https://www.ijcesen.com/index.php/ijcesen/article/view/1021
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Satishkumar Mallappa, Chandrashekar Gudada, P. Megana Santhoshi, Raghavendra (Author)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
