Deep Neural Network with a Characteristic Analysis for Seal Stroke Recognition
Authors: 
Xingyu Cui, Yong Li, Lili XuAuthors Info & Claims
Xingyu Cui, Yong Li, Lili XuAuthors Info & ClaimsArticle No.: 152, Pages 1 - 22
Abstract
Seal characters are derived from ancient Chinese pictographs, naturally inheriting pictographic characteristics and complex structures. As the essential components of seal characters, seal strokes play a vital role in seal character recognition, composition, and writing, so accurate recognition of seal strokes can greatly promote the investigation of seal characters. Inspired by curve fitting, we propose a new model called the characteristic analysis neural network (CANN) for seal stroke recognition. Instead of indiscriminate grasping of feature information in regular neural networks, we design an efficient approximation technique based on the piecewise Bezier curves that can effectively facilitate structural compression and lossless feature extraction. The feature extraction capability of Bezier approximation helps the methodology achieve impressive recognition accuracy not only on the seal strokes but also on any curve-based symbols. Furthermore, the hierarchical structure of the deep learning strategy is inherited and improved for better performance with high generalisation. Experiments conducted on different types of strokes verify that CANN obtains superior performance on both seal strokes and other smooth symbols. The robustness and the effectiveness of CANN are also demonstrated with minimal learning cost compared to other state-of-art models.
References
[1]
I. S. I. Abuhaiba, S. Dattat, and M. J. J. Holt. 1995. Fuzzy state machines to recognize totally unconstructed handwritten strokes. Image. Vis. Comput. 13, 10 (1995), 755–769.
[2]
Alireza Alaei, Partha Pratim Roy, and Umapada Pal. 2016. Logo and seal based administrative document image retrieval: A survey. Comput. Sci. Rev. 22 (2016), 47–63.
[3]
Hussein Almuallim and Shoichiro Yamaguchi. 1987. A method of recognition of Arabic cursive handwriting. IEEE Trans. Pattern Anal. Mach. Intell. PAMI-9, 5 (1987), 715–722.
[4]
Zhenlong Bai and Qiang Huo. 2005. A study on the use of 8-directional features for online handwritten Chinese character recognition. In Proceedings of the 8th International Conference on Document Analysis and Recognition (ICDAR’05). 262–266.
[5]
François Chollet and J. J. Allaire. 2018. Deep Learning with R. Manning Publications. Retrieved from https://livebook.manning.com/book/deep-learning-with-r/
[6]
Dan C. Ciresan, Ueli Meier, and Jürgen Schmidhuber. 2012. Multi-column deep neural networks for image classification. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 3642–3649.
[7]
Xingyu Cui, Yong Li, and Lili Xu. 2023. Adaptive extension fitting scheme: An effective curve approximation method using piecewise Bézier technology. IEEE Access 11 (2023), 58422–58435.
[8]
Janez Demšar. 2006. Statistical comparisons of classifiers over multiple data sets. J. Mach. Learn. Res. 7 (2006), 1–30.
[9]
Bradley Efron and Trevor Hastie. 2016. Computer Age Statistical Inference: Algorithms, Evidence, and Data Science (1st ed.). Cambridge University Press, USA. 351–374.
[10]
Ji Gan, Weiqiang Wang, and Ke Lu. 2019. A new perspective: Recognizing online handwritten Chinese characters via 1-dimensional CNN. Inf. Sci. 478 (2019), 375–390.
[11]
Xavier Glorot and Yoshua Bengio. 2010. Understanding the difficulty of training deep feedforward neural networks. In Proceedings of the 13th International Conference on Artificial Intelligence and Statistics. 249–256.
[12]
Ian Goodfellow, Yoshua Bengio, and Aaron Courville. 2016. Deep Learning. MIT Press. Retrieved from http://www.deeplearningbook.org
[13]
Benjamin Graham. 2013. Sparse arrays of signatures for online character recognition. ArXiv abs/1308.0371 (2013).
[14]
Allan Y. Hasegawa, Roberto S. U. Rosso, and Marcos Sales Guerra Tsuzuki. 2013. Bézier curve fitting with a parallel differential evolution algorithm. IFAC Proc. Vol. 46, 7 (2013), 233–238. DOI:DOI:
[15]
Christopher Martin Holt, Alan James Stewart, Maurice Clint, and Ronald H. Perrott. 1987. An improved parallel thinning algorithm. Commun. ACM 30, 2 (1987), 156–160.
[16]
Baotian Hu, Xin Liu, Xiangping Wu, and Qingcai Chen. 2020. Stroke sequence-dependent deep convolutional neural network for online handwritten Chinese character recognition. IEEE Trans. Neural Netw. Learn. Syst. 31, 11 (2020), 4637–4648.
[17]
Kyung-Won Kang and Jin Hyung Kim. 2004. Utilization of hierarchical, stochastic relationship modeling for Hangul character recognition. IEEE Trans. Pattern Anal. Mach. Intell. 26, 9 (2004), 1185–1196.
[18]
Fumitaka Kimura, Kenji Takashina, Shinji Tsuruoka, and Yasuji Miyake. 1987. Modified quadratic discriminant functions and the application to Chinese character recognition. IEEE Trans. Pattern Anal. Mach. Intell. PAMI-9 (1987), 149–153.
[19]
Songxuan Lai, Lianwen Jin, and Weixin Yang. 2017. Toward high-performance online HCCR: A CNN approach with DropDistortion, path signature and spatial stochastic max-pooling. ArXiv abs/1702.07508 (2017).
[20]
Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner. 1998. Gradient-based learning applied to document recognition. Proc. IEEE 86 (1998), 2278–2324.
[21]
A. D. I. Selevan Lev and Miriam Furs. 1989. Recognition of handwritten Hebrew one-stroke letters by learning syntactic representations of symbols. IEEE Trans. Syst., Man, Cybern. 19, 5 (1989), 1306–1313.
[22]
Yunxin Li, Qian Yang, Qingcai Chen, Baotian Hu, Xiaolong Wang, Yuxin Ding, and Lin Ma. 2023. Fast and robust online handwritten Chinese character recognition with deep spatial and contextual information fusion network. IEEE Trans. Multim. 25 (2023), 2140–2152.
[23]
Chia-Wei Liao and Jun Siung Huang. 1990. Stroke segmentation by bernstein-bezier curve fitting. Pattern Recog. 23 (1990), 475–484.
[24]
Cheng-Lin Liu, Stefan Jäger, and M. Nakagawa. 2004. Online recognition of Chinese characters: The state-of-the-art. IEEE Trans. Pattern Anal. Mach. Intell. 26, 2 (2004), 198–213.
[25]
Cheng-Lin Liu, Fei Yin, Da-Han Wang, and Qiu-Feng Wang. 2011. CASIA online and offline Chinese handwriting databases. In Proceedings of the International Conference on Document Analysis and Recognition. 37–41.
[26]
Cheng-Lin Liu, Fei Yin, Da-Han Wang, and Qiu-Feng Wang. 2013. Online and offline handwritten Chinese character recognition: Benchmarking on new databases. Pattern Recog. 46 (2013), 155–162.
[27]
Cheng-Lin Liu, Fei Yin, Qiu-Feng Wang, and Da-Han Wang. 2011. ICDAR 2011 Chinese handwriting recognition competition. In Proceedings of the International Conference on Document Analysis and Recognition. 1464–1469.
[28]
Minting Lu. 2019. The evolution of chinese character writing elements: Focusing on small seal script in shuowen. phdthesis. Beijing: Beijing Normal University, 20–40.
[29]
Asif Masood, Muhammad I. Sarfraz, and Shaiq A. Haq. 2005. Curve approximation with quadratic B-splines. In Proceedings of the 9th International Conference on Information Visualisation (IV’05). 419–424.
[30]
A. T. Mckay and Egon S. Pearson. 1933. ii) A note on the distribution of range in samples of n. Biometrika 25 (1933), 415–420.
[31]
Michael E. Mortenson. 1999. Mathematics for Computer Graphics Applications. Industrial Press Inc. 264 pages. 99010096
[32]
P. Nemenyi. 1963. Distribution-free Multiple Comparisons. Princeton University.
[33]
Priza Pandunata and Siti Mariyam H. J. Shamsuddin. 2010. Differential evolution optimization for Bezier curve fitting. DOI:DOI:
[34]
Les Piegl and Wayne Tiller. 1996. The NURBS Book. Springer Berlin. 410 pages. 95032273
[35]
Hartmut Prautzsch, Wolfgang Boehm, and Marco Paluszny. 2002. Bézier and B-Spline Techniques (1st ed.). Springer-Verlag Berlin.
[36]
Kaushik Roy. 2012. Stroke-database design for online handwriting recognition in Bangla. Int. J. Mod. Eng. Res. 2, 4 (2012), 2534–2540.
[37]
Muhammad Sarfraz. 2008. Interactive Curve Modeling. Springer London. 173–194.
[38]
Muhammad I. Sarfraz and Syed Arshad Raza. 2002. Visualization of Data Using Genetic Algorithm. Springer London. 535–544. DOI:DOI:
[39]
Karen Simonyan and Andrew Zisserman. 2015. Very deep convolutional networks for large-scale image recognition. CoRR abs/1409.1556 (2015).
[40]
Sukhdeep Singh, Anuj Sharma, and Indu Chhabra. 2016. Online handwritten Gurmukhi strokes dataset based on minimal set of words. ACM Trans. Asian Low-resour. Lang. Inf. Process. 16, 1 (2016), 1–20.
[41]
Nouri A. Suleiman. 2013. Least squares data fitting with quadratic Bezier curves. Proc. YSU, Phys. Math. Sci.2 (2013), 42–49.
[42]
Bin Sun, Shaojun Hua, Shutao Li, and Jun Sun. 2019. Graph-matching-based character recognition for Chinese seal images. Sci. China Inf. Sci. 62, Sept. (2019), 1–14.
[43]
Xiaokun Sun and Jifeng Huang. 2018. A thinning method for international phonetic alphabet characters. J. Graph. 39, 2 (2018), 214–220.
[44]
Junjian Tang and Jun Guo. 2018. A new method for stroke order recognition of handwritten Chinese characters. In Proceedings of the the 2nd International Conference on Video and Image Processing. 7–11.
[45]
Edson K. Ueda, André Kubagawa Sato, Thiago de C. Martins, Rogério Y. Takimoto, Roberto Silvio Ubertino Rosso, and Marcos Sales Guerra Tsuzuki. 2020. Curve approximation by adaptive neighborhood simulated annealing and piecewise Bézier curves. Soft Comput. 24 (2020), 18821–18839.
[46]
Edson K. Ueda, Marcos S. G. Tsuzuki, and Ahmad Barari. 2018. Piecewise Bézier curve fitting of a point cloud boundary by simulated annealing. In Proceedings of the 13th IEEE International Conference on Industry Applications (INDUSCON’18). 1335–1340.
[47]
Ning Wang. 2015. Introduction to chinese character configuration. The Commercial Press, 77–79.
[48]
Shen Xu. Eastern Han100-121. Shuo Wen Jie Zi. China Publishing House. https://szsw.bnu.edu.cn/book/org-page.action?page=7&num=1&zi=%E4%B8%80
[49]
Weixin Yang, Lianwen Jin, Dacheng Tao, Zecheng Xie, and Ziyong Feng. 2015. DropSample: A new training method to enhance deep convolutional neural networks for large-scale unconstrained handwritten Chinese character recognition. Pattern Recog. 58 (2015), 190–203.
[50]
Fei Yin, Qiu-Feng Wang, Xu-Yao Zhang, and Cheng-Lin Liu. 2013. ICDAR 2013 Chinese handwriting recognition competition. In Proceedings of the 12th International Conference on Document Analysis and Recognition. 1464–1470.
[51]
Jia Zeng and Zhi Qiang Liu. 2008. Markov random field-based statistical character structure modeling for handwritten Chinese character recognition. IEEE Trans. Pattern Anal. Mach. Intell. 30, 5 (2008), 767–780.
[52]
T. Y. Zhang and Ching Yee Suen. 1984. A fast parallel algorithm for thinning digital patterns. Commun. ACM 27, 3 (1984), 236–239.
[53]
Xu-Yao Zhang, Yoshua Bengio, and Cheng-Lin Liu. 2016. Online and offline handwritten Chinese character recognition: A comprehensive study and new benchmark. Pattern Recog. 61 (2016), 348–360.
[54]
Qing Zhao and Yingmin Tang. 2009. Shape recognition-based approach to Chinese character strokes’ classification. Comput. Technol. Devel. 19, 10 (2009), 14–17.
[55]
Zhuoyao Zhong, Lianwen Jin, and Zecheng Xie. 2015. High performance offline handwritten Chinese character recognition using GoogLeNet and directional feature maps. In Proceedings of the 13th International Conference on Document Analysis and Recognition (ICDAR’15). 846–850.
[56]
Xiaowen Zhou and Yong Li. 2000. Seal character front library (computer truetype font library and input method). Beijing Publishing House, Beijing.
[57]
D. Álvarez, R. Fernández, and L. Sánchez. 2015. Stroke-based intelligent character recognition using a deterministic finite automaton. Logic J. IGPL 23, 3 (2015), 463–471.
Index Terms
- Deep Neural Network with a Characteristic Analysis for Seal Stroke Recognition
Recommendations
Handwritten Character Recognition using Deep Neural Networks
DSMLAI '21': Proceedings of the International Conference on Data Science, Machine Learning and Artificial IntelligenceThe preliminary work performed in this manuscript is to recognize Handwritten English Characters using a multilayer perceptron. The standard EMNIST dataset of handwritten English characters is used here. The preprocessing of images included Binarization ...
Scanning Neural Network for Text Line Recognition
DAS '12: Proceedings of the 2012 10th IAPR International Workshop on Document Analysis SystemsOptical character recognition (OCR) of machine printed Latin script documents is ubiquitously claimed as a solved problem. However, error free OCR of degraded or noisy text is still challenging for modern OCR systems. Most recent approaches perform ...
UrduDeepNet: offline handwritten Urdu character recognition using deep neural network
AbstractHandwritten Urdu character recognition system faces several challenges including the writer-dependent variations and non-availability of benchmark databases for cursive writing scripts. In this study, we propose a handwritten Urdu character ...
Comments
Information & Contributors
Information
Published In
November 2024
248 pages
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 21 November 2024
Online AM: 30 July 2024
Accepted: 25 May 2024
Revised: 16 July 2023
Received: 30 October 2021
Published in TALLIP Volume 23, Issue 11
Check for updates
Author Tags
Qualifiers
- Research-article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 129Total Downloads
- Downloads (Last 12 months)129
- Downloads (Last 6 weeks)5
Reflects downloads up to 02 Mar 2025
Other Metrics
Citations
View Options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in