Optimised CNN Architectures for Handwritten Arabic Character Recognition
Source abstract
Abstract
Handwritten character recognition is considered challenging compared with machine-printed characters due to the different human writing styles. Arabic is morphologically rich, and its characters have a high similarity. The Arabic language includes 28 characters. Each character has up to four shapes according to its location in the word (at the beginning, middle, end, and isolated). This paper proposed 12 CNN architectures for recognizing handwritten Arabic characters. The proposed architectures were derived from the popular CNN architectures, such as VGG, ResNet, and Inception, to make them applicable to recognizing character-size images. The experimental results on three well-known datasets showed that the proposed architectures significantly enhanced the recognition rate compared to the baseline models. The experiments showed that data augmentation improved the models’ accuracies on all tested datasets. The proposed model outperformed most of the existing approaches. The best achieved results were 93.05%, 98.30%, and 96.88% on the HIJJA, AHCD, and AIA9K datasets.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Handwritten Text Recognition Techniques · Hand Gesture Recognition Systems · Vehicle License Plate Recognition
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (49)
Show source-provided references
1. A Survey of OCR in Arabic language: Applications, techniques, and challenges10.3390/app13074584 ↗
2. OCR-MRD: Performance analysis of different optical character recognition engines for medical report digitization10.1007/s41870-023-01610-2 ↗
3. Automated invoice data extraction using image processing
4. Automated player identification and indexing using two-stage deep learning network10.1038/s41598-023-36657-5 ↗
5. The system for efficient indexing and search in the large archives of scanned historical documents
6. OCR based cheque validation using image processing
7. Automatic imagery bank cheque data extraction based on machine learning approaches: A comprehensive survey10.1007/s11042-023-14534-7 ↗
8. Text region localization and recognition for ID card identification using deep learning approaches
9. C. T. Haile, “Customers identity card data detection and recognition using image processing,” Ph.D. dissertation, Dept. of Computer Science, St. Mary’s Univ., Addis Ababa, Ethiopia, 2023.
10. OCR assisted translator
11. Optical character recognition and neural machine translation using deep learning techniques
12. Handwritten character recognition models based on convolutional neural networks
13. A. Bochkovskiy, C. Y. Wang, and H. Y. M. Liao, “YOLOv4: Optimal speed and accuracy of object detection,” arXiv preprint arXiv:2004.10934, Apr. 2020.
14. Faster R-CNN: Towards real-time object detection with region proposal networks10.1109/TPAMI.2016.2577031 ↗
15. SSD: Single shot MultiBox detector
16. K. Simonyan and Z. Andrew, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556, 2014.
17. Going deeper with convolutions
18. Deep residual learning for image recognition
19. EfficientNet: Rethinking model scaling for convolutional neural networks
20. A. Dosovitskiy et al., “An image is worth 16×16 words: Transformers for image recognition at scale,” arXiv preprint arXiv:2010.11929, 2010.
21. Z. Andrew et al., “MobileNets: Efficient convolutional neural networks for mobile vision applications,” arXiv preprint arXiv:1704.04861, 2017.
22. C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethinking the inception architecture for computer vision,” arXiv preprint arXiv:1512.00567, 2015.10.1109/CVPR.2016.308 ↗
23. ImageNet classification with deep convolutional neural networks10.1145/3065386 ↗
24. Gradient-based learning applied to document recognition10.1109/5.726791 ↗
25. An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition10.1109/TPAMI.2016.2646371 ↗
26. The World Bank, “Arab world population,” Accessed: Dec. 24, 2023. [Online]. Available: https://data.worldbank.org.cn/
27. Arabic optical character recognition: A review
28. A printed arabic optical character recognition system using deep learning10.3844/jcssp.2022.1038.1050 ↗
29. Robust language-independent OCR system
30. Arabic OCR evaluation tool
31. Artificially intelligent readers: An adaptive framework for original handwritten numerical digits recognition with OCR methods10.3390/info14060305 ↗
32. Arabic handwritten characters recognition using convolutional neural network
33. Arabic handwritten character recognition based on convolution neural networks and support vector machine
34. An effective deep learning approach for improving off-line arabic handwritten character recognition
35. Recognizing arabic handwritten characters using deep learning and genetic algorithms10.1007/s11042-021-11185-4 ↗
36. Improved Arabic alphabet characters classification using convolutional neural networks (CNN)10.1155/2022/9965426 ↗
37. Optimization of hijaiyah letter handwriting recognition model based on deep learning
38. Arabic calligraphy image analysis with transfer learning
39. An ensemble learning approach using decision fusion for the recognition of Arabic handwritten characters
40. Combination of machine learning algorithms and ResNet50 for Arabic handwritten classification
41. Hybrid Arabic handwritten character segmentation using CNN and graph theory algorithm
42. M. Abadi et al., “TensorFlow: Large-scale machine learning on heterogeneous distributed systems,” arXiv preprintarXiv:1603.04467, Mar. 2016.
43. Arabic handwriting recognition system using convolutional neural network10.1007/s00521-020-05070-8 ↗
44. M. Torki, M. E. Hussein, A. Elsallamy, M. Fayyaz, and S. Yaser, “Window-based descriptors for Arabic handwritten alphabet recognition: A comparative study on a novel dataset,” arXiv preprint arXiv:1411.3519, 2014.
45. Arabic hand-written character recognition based on deep convolutional neural networks
46. Arabic handwritten recognition based on deep convolutional neural network
47. A new Arabic handwritten character recognition deep learning system10.1007/s00521-020-05397-2 ↗
48. Handwritten Arabic character recognition: Comparison of conventional machine learning and deep learning approaches
49. An intelligent approach for Arabic handwritten letter recognition using convolutional neural network10.7717/peerj-cs.995 ↗