فراتفکیک‌پذیری مبتنی بر نمونه تک‌تصویر متن با روش نزول گرادیان ناهمزمان ترتیبی
محورهای موضوعی : مهندسی برق و کامپیوتر
علی عابدی
1
,
احسان‌اله کبیر
2
1 - دانشگاه تربیت مدرس
2 - دانشگاه تربیت مدرس
کلید واژه: بهسازی تصویر متن افزایش تفکیک‌پذیری فراتفکیک‌پذیری بیزی فراتفکیک‌پذیری مبتنی بر نمونه الگوریتم نزول گرادیان,
چکیده مقاله :
در اين مقاله، یک روش جدید برای افزایش تفکیک‌پذیری تک‌تصویری تصاویر متن ارائه می‌شود. این روش مبتنی بر نمونه است یعنی برای فراتفکیک‌پذیری از یک مجموعه نمونه آموزشی که شامل وصله‌های با تفکیک‌پذیری بالا و پایین است استفاده می‌شود. بر اساس قاعده بیزی، یک تابع به عنوان درست‌نمایی و سه تابع به عنوان دانش اولیه در نظر گرفته می‌شوند. تابع مربوط به درست‌نمایی میزان شباهت با تصویر اولیه را توصیف می‌کند. سه تابع مربوط به دانش اولیه خصوصیات دومُدی بودن تصویر متن، یکنواخت‌بودن نواحی پس‌زمینه و متن و نزدیک‌بودن به مجموعه نمونه آموزشی را توصیف می‌کنند. با کمینه‌کردن این توابع انرژی طی فرایند تکرارشونده نزول گرادیان ناهمزمان ترتیبی، تصویر با تفکیک‌پذیری بالا به دست می‌آید. به جای کمینه‌کردن هم‌زمان ترکیب خطی توابع، آنها به ترتیب و با توجه به این که در تکرارهای متوالی الگوریتم چه تغییراتی در تصویر متن رخ می‌دهد کمینه می‌گردند. به این ترتیب دیگر نیازی به تعیین ضرایب ترکیب خطی توابع که برای تصاویر مختلف متغیر هستند نخواهد بود. نتایج آزمایش‌ها روی بیست تصویر متن با قلم‌ها، تفکیک‌پذیری‌ها، تارشدگی‌ها و نویزهای مختلف عملکرد بهتر و با حجم محاسباتی کمتر روش ارائه‌شده نسبت به روش‌های مشابه قبلی را نشان می‌دهد.
In this paper, a new method for resolution enhancement of single document images is presented. The proposed method is example based using an example set of low-resolution and high-resolution training patches. According to the Bayes rule, one function is considered as the likelihood or data-fidelity term that measures the fidelity of the output high-resolution to the input low-resolution image. As well, three other functions are considered as the regularization terms containing the prior knowledge about the desired high-resolution document image. Three priors which are fulfilled by the regularization terms are bimodality of document images, smoothness of background and text regions, and similarity to the patches in the example set. By minimizing these four energy functions through the iterative procedure of asynchronous sequential gradient descent, the HR image is reconstructed. Instead of synchronous minimization of the linear combination of these functions, they are minimized in order and according to the gradual changes in their values and in the updating HR image. Therefore, determining the coefficients of the linear combination, which are variable for input images, is no longer required. In the experimental results on twenty document images with different fonts, at different resolutions, and with different amounts of noise and blurriness, the proposed method achieves significant improvements in visual image quality and in reducing the computational complexity.
[1] A. Abedi and E. Kabir, "Stroke width-based directional total variation regularisation for document image super resolution," IET Image Processing, vol. 10, no. 2, pp. 158-166, Feb. 2016.
[2] P. Milanfar, Super-Resolution Imaging, vol. 1, CRC Press, 2010.
[3] K. Donaldson and G. Myers, "Bayesian super-resolution of text in video with a text-specific bimodal prior," Int. J. Document Anal. Recognit., vol. 7, no. 2, pp. 159-167, Jul. 2005.
[4] C. M. Thillou and M. Mirmehdi, "An introduction to super-resolution text," Digital Document Processing, Advances in Pattern Recognition, vol. 16, no. 17, pp. 305-327, Sep. 2007.
[5] D. Datsenko and M. Elad, "Example-based single document image super-resolution: a global MAP approach with outlier rejection," Multidim Syst Sign Process, vol. 18, no. 2, pp. 103-121, Sep. 2007.
[6] M. Elad and D. Datsenko, "Example-based regularization deployed to super-resolution reconstruction of a single image," The Computer J., vol. 52, no. 1, pp. 15-30, Oct. 2009.
[7] J. Park, Y. Kwon, and J. H. Kim, "An example-based prior model for text image super-resolution," in Proc. 8th Int. Conf. on Document Analysis and Recognition, vol. 1, pp. 374-378, Sep. 2005.
[8] R. Zeyde, M. Elad, and M. Protter, "On single image scale-up using sparse-representations," Curves and Surfaces Lecture Notes in Computer Science, vol. 6920, pp. 711-730, Jan. 2012.
[9] R. Walha, F. Driram, F. Lebourgeois, and A. M. Alimi, "Super-resolution of single text image by sparse representation," in Proc. of the Workshop on Document Analysis and Recognition DAR'12, pp. 22-29, Aug. 2012.
[10] G. Caner and I. Haritaoglu, "ShapeDNA: effective character restoration and enhancement for arabic text documents," in Proc. 20th Int. Conf. on Pattern Recognition, ICPR'10, pp. 2053-2056, Jul. 2010.
[11] S. Baker and T. Kanade, "Limits on super-resolution and how to break them," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 24, no. 9, pp. 1167-1183, Nov. 2002.
[12] W. T. Freeman, E. C. Pasztor, and O. T. Carmichael, "Learning low-level vision," International J. of Computer Vision, vol. 40, no. 1, pp. 25-47, Oct. 2000.
[13] P. Thouin and C. Chang, "A method for restoration of low-resolution document images," Int. J. Document Anal. Recognit., vol. 2, no. 4, pp. 200-210, Jun. 2000.
[14] H. Q. Luong and W. Philips, "Robust reconstruction of low-resolution document images by exploiting repetitive character behaviour," International J. on Document Analysis and Recognition, vol. 11, no. 1, pp. 39-51, Oct. 2008.
[15] J. Banerjee and C. V. Jawahar, "Super-resolution of text images using edge-directed tangent field," in Proc. 8th IAPR Int. Workshop on Document Analysis Systems, DAS'08, pp. 76-83, Nov. 2008.
[16] A. Kheradmand and P. Milanfar, "A general framework for regularized, similarity-based image restoration," IEEE Trans. on Image Processing, vol. 23, no. 12, pp. 5136-5151, Dec. 2014.
[17] J. H. Friedman, J. L. Bentley, and R. A. Finkel, "An algorithm for finding best matches in logarithmic expected time," ACM Trans. on Mathematical Software, vol. 3, no. 3, pp. 209-226, Feb. 1977.
[18] M. V. W. Zibetti, F. S. V. Bazan, and J. Mayer, "Determining the regularization parameters for super-resolution problems," Signal Processing, vol. 88, no. 12, pp. 2890-2901, Dec. 2008.
[19] A. Panagiotopoulou and V. Anastassopoulos, "Super-resolution image reconstruction techniques: trade-offs between the data-fidelity and regularization terms," Information Fusion, vol. 13, no. 3, pp. 185-195, Jul. 2012.
[20] A. Agarwal and J. C. Duchi, Distributed Delayed Stochastic Optimization, arXiv: 1104.5525, 2011.
[21] H. Lu, A. Kot, and Y. Shi, "Distance-reciprocal distortion measure for binary document images," IEEE Signal Processing Letters, vol. 11, no. 2, pp. 228-231, Feb. 2004.
[22] S. M. Pincus, I. M. Gladstone, and R. A. Ehrenkranz, "A regularity statistic for medical data analysis," J. of Clinical Monitoring and Computing, vol. 7, no. 4, pp. 335-345, Feb. 1991.
[23] G. Louloudis, B. Gatos, I. Pratikakis, and C. Halatsis, "Text line detection in handwritten documents," Pattern Recognition, vol. 41, no. 12, pp. 3758-3772, Dec. 2008.
[24] L. Zheng, S. Wang, and Y. Liu, "Information theoretic regularization for semi-supervised boosting," in Proc. of the 15th ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, KDD'09, pp. 1017-1026, Aug. 2009.
[25] L. M. Lorigo and V. Govindaraju, "Offline Arabic handwriting recognition: a survey," IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 28, no. 5, pp. 712-724, Mar. 2006.
