GPU Approach for Handwritten Devanagari Document Binarization
Abstract
The optical character recognition (OCR) is the process of converting scanned images of machine printed or handwritten text, numerals, letters, and symbols into a computer processable format such as ASCII. For creating OCR’s paperless application, a system of high speed and of better accuracy is required. Parallelization of algorithm using graphics processing unit (GPU) along with CPU can be used to speed up the processing. In GPU computing, the compute-intensive operations are performed on GPU while serial code still runs on CPU. Binarization is one of the most fundamental preprocessing techniques in the area of image processing and pattern recognition. This paper proposes an adaptive threshold binarization algorithm for GPU. The aim of this research work is to speed up binarization process that eventually will help to accelerate the processing of document recognition. The algorithm implementation is done using Compute Unified Device Architecture (CUDA) software interface by NVIDIA. An average speedup of 2× is achieved on GPU GeForce 210 having 16 CUDA cores and 1.2 compute level, over the serial implementation.
Keywords
CUDA GPU OCR Binarization Parallelization Pattern recognitionReferences
- 1.I.K. Sethi and B. Chatterjee, “Machine Recognition of constrained Handprinted Devnagari”, Pattern Recognition, Vol. 9, pp. 69–75, 1977.Google Scholar
- 2.Pal, U. and Chaudhuri, B.B., “Indian script character recognition: a survey”. Pattern Recognition 37, 1887–1899, 2004.Google Scholar
- 3.S. Arora, “Studies on some Soft Computing Techniques: A Case Study for Constrained Handwritten Devnagari Characters and Numerals”, International Journal of Scientific & Engineering Research, Volume 4, Issue 7, 2013.Google Scholar
- 4.Xiu, P., Baird, H.S.: Whole-book recognition. IEEE Trans. Pattern Anal. Mach. Intell. 34(12), 2467–2480, 2012.Google Scholar
- 5.Kae, A., Huang, G.B., Doersch, C., Learned-Miller, E.G., “Improving state-of-the-art OCR through high-precision document-specific modeling” In: CVPR, pp. 1935–1942,2010.Google Scholar
- 6.Eikvil, L. “OCR-Optical Character Recognition”, 1993, http://www.nr.no/*eikvil/OCR.pdf.
- 7.Gaceb, D., Eglin, V., Lebourgeois, F. “A new mixed binarization method used in real time application of automatic business document and postal mail sorting” Int. Arab J. Inf. Technol. 10(2), 179–188, 2013.Google Scholar
- 8.M. Hanmandlu and O.V. Ramana Murthy, “Fuzzy Model Based Recognition of Handwritten Hindi Numerals”, In Proc. Intn Conference on Cognition and Recognition, 490–496, 05.Google Scholar
- 9.M. Hanmandlu, O.V. Ramana Murthy, Vamsi Krishna Madasu, “Fuzzy Model based recognition of Handwritten Hindi characters”, IEEE Computer society, Digital Image Computing Techniques and Applications, 2007.Google Scholar
- 10.Reena Bajaj, Lipika Dey, and S. Chaudhury, “Devnagari numeral recognition by combining decision of multiple connectionist classifiers”, Sadhana, Vol. 27, part. 1, pp.-59–72, 02.Google Scholar
- 11.S. Kumar, C. Singh, “A Study of Zernike Moments and its use in Devnagari Handwritten Character Recognition”, Intl. Conf. On Cognition and Recognition, 514–520, 05.Google Scholar
- 12.U. Bhattacharya, B. B. Chaudhuri, R. Ghosh and M. Ghosh, “On Recognition of Handwritten Devnagari Numerals”, In Proc. of the Workshop on Learning Algorithms for Pattern Recognition, Sydney, pp. 1–7, 2005.Google Scholar
- 13.N. Sharma, U. Pal, F. Kimura, S. pal, “Recognition of Off Line Handwritten Devnagari Characters using Quadratic Classifier”, ICCGIP 2006, LNCS 4338, pp 805–816, 2006.Google Scholar
- 14.S. Arora, D. Bhattacharya, M. Nasipuri, “ Recognition of Non-Compound Handwritten Devnagari Characters using a Combination of MLP and Minimum Edit Distance”, International Journal of Computer Science and Security (IJCSS), Volume (4): Issue (1), 2010.Google Scholar
- 15.Brij Mohan Singh, Rahul Sharma, Ankush Mittal, Debashish Ghosh, “Parallel Implementation of Otsu’s Binarization Approach on GPU”, International Journal of Computer Applications (0975 – 8887) Volume 32–No. 2, October 2011.Google Scholar
- 16.Sauvola, J. and Pietikainen, M., “Adaptive document image binarization” Pattern Recognition, Vol. 33, 225–236, 2000.Google Scholar
- 17.Kim, I.K., Jung, D.W. and Park, R.H., “Document image binarization based on topographic analysis using a water flow model”, Pattern Recognition, Vol. 35, 265–277, 2002.Google Scholar
- 18.Gatos, B., Pratikakis, I. and Perantonis, S. J. “Adaptive degraded document image binarization”, Pattern Recognition, Vol. 39, 317–327, 2006.Google Scholar
- 19.Chang, Y.F., Pai, Y.T. and Ruan, S.J. “An efficient thresholding algorithm for degraded document images based on intelligent block detection”, IEEE Intern Conf on Systems, Man, and Cybernetics, 667–672, 08.Google Scholar
- 20.Valizadeh, M., Komeili, M., Armanfard, N. and Kabir, E., “Degraded document image binarization based on combination of two complementary algorithms”, International Conf. of Advances in Computational Tools for Engineering Applications, IEEE, 595–599, 2009.Google Scholar
- 21.Owens, J. D., Luebke, D., Govindaraju, N., Harris, M., Kruger, J., Lefohn, A. E. and Purcell, T. J., “A survey of general-purpose computation on graphics hardware”, In proceeding of Eurographics, State of the Art Reports, 21–51, 2005.Google Scholar
- 22.M. Soua, R. Kachouri, M. Akil, “ GPU parallel implementation of the new hybrid binarization based on Kmeans method HBK”, Journal of Real-Time Image Processing, 14.Google Scholar
- 23.Yi, Y., Lai, C., Petrov, S.: Efficient parallel CKY parsing using GPUs. J. Log. Comput. 24(2), 375–393, 2014.Google Scholar
- 24.George Cherian Panappally, M.S. Dhanesh, “Design of graphics processing unit for image processing”, 1st Intern Conf on Computational Systems & Communications, 2014.Google Scholar
- 25.Larsen, E. S., McAllister, D.,” Fast Matrix Multiplies using Graphics Hardware”, International Conference for High Performance Computing and Communications, 159–168, 2001.Google Scholar
- 26.Trendall C. and Stewart, A. J., “General calculations using graphics hardware with applications to interactive caustics Rendering Techniques”, 11th Eurographics Workshop on Rendering, 287–298, 2000.Google Scholar
- 27.Li, Wei, Xiaoming, A. & Kaufman, “Implementing lattice boltzmann computation on graphics h/w,” Intern. Conf. for High Performance Computing and Communins, 2001.Google Scholar
- 28.Kruger, J. and Westermann, R., “Linear operators for GPU implementation of numerical algorithms”, In Proceedings of SIGGRAPH, San Diego, 908–916, 2003.Google Scholar
- 29.Mizukami, Y., Koga, K. and Torioka, T., “A handwritten character recognition system using hierarchical extraction of displacement”, IEICE, J77-D-II(12): 2390–2393, 1994.Google Scholar
- 30.Steinkraus, D., Buck, I., and Simard, P. Y., “GPUs for machine learning algorithms”, International Conference of Document Analysis and Recognition, 1115–1120, 2005.Google Scholar
- 31.Ilie, “A. Optical character recognition on graphics hardware”, www.cs.unc.edu/~adyilie/IP/Final.pdf.
- 32.Lazzara, G., Graud, T., “Efficient multiscale Sauvola’s binarization”, Int. Journal of Document Analysis and Recognition, 2(14), 105–123,2014.Google Scholar
- 33.Oh, K.S. and Jung, K., “GPU implementation of neural networks”, Pattern Recognition, Elsevier, 1311–1314, 2004.Google Scholar
- 34.Jung, K. “Neural Network-based text localization in color image”,. Pattern Recognition Letters, Vol. 22, (4), 1503–1515, 2001.Google Scholar
- 35.Singh, B.M., Mittal A. and Ghosh, D., “Parallel implementation of Devanagari text line and word segmentation approach on GPU”, IJCA 24(9): 7–14, 2011.Google Scholar
- 36.NVIDIA Corporation: NVIDIA CUDA programming guide. Jan 2007, http://developer.download.nvidia.com/compute/cuda/2_0/docs/NVIDIA_CUDA_Programming_Guide_2.0.pdf.