Abstract
Automatic scene text recognition is an interesting problem in computer vision and Internet of things. It may facilitate intelligent interaction between machines and mankind in today’s cloud-enabled civilization. In this paper, we present a method for dissected scene character recognition. At first, color images are converted into grayscale and then some noise removal and pre-processing operations are applied. Next, we normalize them to bring them to a uniform dimension and compute features for training and prediction. Experimenting on scene characters at three different levels of complexities i.e. relatively good images, relatively bad images, and combined images with multiple classifiers such as naïve Bayes, KNN, MLP, random forest and SVM, detail results are reported. Highest accuracies i.e. 74.48% for good images only, 59.13% on bad images only and 71.52% for overall images, are obtained with the SVM classifier. Comparison with similar state-of-the-art methods is also included and our method is found to outperform others.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Lin H, Yang P, Zhang F (2020) Review of scene text detection and recognition. Arch Comput Meth Eng 27(2):433–454
Zhu Y, Yao C, Bai X (2016) Scene text detection and recognition: recent advances and future trends. Front Comput Sci 10(1):19–36
Sengupta P, Mollah AF (2020) Journey of scene text components recognition: progress and open issues. Multimedia Tools Appl. Springer, 80(4):6079–6104
Liu X, Meng G, Pan C (2019) Scene text detection and recognition with advances in deep learning: a survey. Int J Doc Anal Recogn 22(2):143–162
Karatzas D, Shafait F, Uchida S, Iwamura M (2013) ICDAR 2013 robust reading competition. In: Proceedings of 12th international conference on document analysis and recognition. IEEE, pp 1484–1493
Iwamura M (2018) Adv Scene Text Datasets: arXiv preprint arXiv:1812.05219
Lucas S, Panaretos M, Sosa AL, Tang A Wong, S and Young R (2003) ICDAR 2003 robust reading competitions. In: Proceedings of seventh international conference on document analysis and recognition. IEEE, pp 682–687
Karatzas D, Gomez-Bigorda L, Nicolaou A, Ghosh S, Bagdanov A, Iwamura M, Matas J, Neumann L, Chandrasekhar VR, Lu S, Shafait F (2015) ICDAR 2015 competition on robust reading. In: Proceedings of 13th international conference on document analysis and recognition. IEEE, pp 1156–1160
De Campos TE, Babu BR, Varma M (2009) Character recognition in natural images. In: Proceedings of international conference on computer vision theory and application, pp 273–280
Neumann L, Matas J (2010) A method for text localization and recognition in real-world images. In: Proceedings of Asian conference on computer vision. Springer, pp 770–783
Chekol B, Celebi N, Taşci T (2019) Segmented character recognition using curvature based global image feature. Turkish J Electric Eng Comput Sci 27(5):3804–3814
Bai X, Yao C, Liu W (2016) Strokelets: a learned multi-scale mid-level representation for scene text recognition. IEEE Trans Image Proc 25(6):2789–2802
Lin JH, Lazarow J, Yang A, Hong D, Gupta R, Tu Z (2020) Local binary pattern networks. In: Proceedings of IEEE winter conference on applications of computer vision, pp 825–834
Sundin H, Josefsson J (2020) Evaluating synthetic training data for character recognition in natural images. Degree Project of KTH Royal Institute of Technology, Sweden
Abdali AR, Ghani RF (2019) Robust character recognition for optical and natural images using deep learning. In: Proceedings of IEEE student conference on research and development, pp 152–156
Barnouti NH, Abomaali M, Al- MHN (2018) An efficient character recognition technique using K-nearest neighbour classifier. Int J Eng Technol 7(4):3148–3153
Jaderberg M, Simonyan K, Vedaldi A, Zisserman A (2014) Deep structured output learning for unconstrained text recognition. In: Proceedings of international conference on learning representations, pp 1–10
Mollah AF, Basu S, Nasipuri M (2012) Computationally efficient implementation of convolution-based locally adaptive binarization techniques. In: Proceedings of international conference on information processing. Springer, pp 159–168
Bapu J, Florinabel DJ (2020) Real-time image processing method to implement object detection and classification for remote sensing images. In: Proceedings of international conference on earth science informatics. Springer, pp 1–13
Goyal V, Shukla A (2020) An enhancement of underwater images based on contrast restricted adaptive histogram equalization for image enhancement. In: Smart innovations in communication and computational sciences. Springer, pp 275–285
Mollah AF, Basu S, Nasipuri M, Basu DK (2013) Handheld mobile device based text region extraction and binarization of image embedded text documents. J Intell Syst 22(1):25–47
Gogna A, Majumdar A (2019) Discriminative autoencoder for feature extraction: application to character recognition. Neural Proc Lett 49(3):1723–1735
Sengupta P, Mollah AF (2020) Scene character recognition with morphological filtering and HOG features. In: Proceedings of an international conference on computing & communication. Springer, pp 1–9
Bakas J, Mahalat MH, Mollah AF (2016) A comparative study of various classifiers for character recognition on multi-script databases. Int J Comput Appl 155(3):1–5
Mollah AF, Basu S, Nasipuri M (2018) An automatic annotation scheme for scene text archival applications. In: Proceedings of international conference on advances in computing and data sciences. Springer, pp 66–76
Acknowledgements
The authors are thankful to the Department of Computer Science and Engineering of Aliah University, Kolkata, India, for providing every kind of support for carrying out this research work. P. Sengupta is further grateful to Dept. of MA & ME, Govt. of West Bengal for providing Swami Vivekananda Merit cum Means Fellowship.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Sengupta, P., Mollah, A.F. (2022). Dissected Scene Character Recognition Using HOG Descriptors. In: Dahal, K., Giri, D., Neogy, S., Dutta, S., Kumar, S. (eds) Internet of Things and Its Applications. Lecture Notes in Electrical Engineering, vol 825. Springer, Singapore. https://doi.org/10.1007/978-981-16-7637-6_18
Download citation
DOI: https://doi.org/10.1007/978-981-16-7637-6_18
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-7636-9
Online ISBN: 978-981-16-7637-6
eBook Packages: EngineeringEngineering (R0)