Abstract
This article considers a method of improving print quality for screenshots. The proposed method is based on detecting and vectorizing text areas on raster images. The main study is dedicated to smooth screen-text segmentation, determining background and text color, improving resolution, and recovering the contour of symbols and approximating them with Bezier curves. The proposed method is resistant to different colors, text sizes, and languages and makes it possible to obtain a sharp and correct text display for printing.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
I. V. Kurilin, I. V. Safonov, M. N. Rychagov, Hokeun Lee, SangHo Kim, and Donchul Choi, “Generation of PDF with vector symbols from scanned document,” Proc. SPIE 8653, Image Quality and System Performance X, 86530R (2013).
A. Coates, B. Carpenter, C. Case, S. Satheesh, B. Suresh, T. Wang, D. J. Wu, and A. Y. Ng, “Text detection and character recognition in scene images with unsupervised feature learning,” in Proc. Int. Conf. on Document Analysis and Recognition (Beijing, 2011).
Farshideh Einsele-Aazami, “Recognition of ultra low resolution, anti-aliased text with small font sizes,” Thesis (Dep. of Computer Science, Univ. of Fribourg, 2008).
S. Wachenfeld, H. Klein, S. Fleischer, and X. Jiang, “Segmentation of very low resolution screen-rendered text,” in Proc. Int. Conf. on Document Analysis and Recognition (Curitiba, 2007), pp. 1153–1157.
S. Gleichman, B. Ophir, A. Geva, M. Marder, E. Barkan, and E. Packer, “Detection and segmentation of antialiased text in screen images,” in Proc. Int. Conf. Document Analysis and Recognition (Beijing, 2011), pp. 424–428.
A. M. Vil’kin, I. V. Safonov, and M. A. Egorova, “Bottom-up document segmentation method based on textural features,” Pattern Recogn. Image Anal. 21 (3), 565–568 (2011).
Y. Freund and R. Schapire, “Experiments with a new boosting algorithm,” in Proc. Int. Conf. on Machine Learning (ICM) (Ames, 1996), pp. 148–156.
A. Vezhnevets and V. Vezhnevets, “Modest AdaBoost–teaching AdaBoost to generalize better,” in Proc. Graphicon (Novosibirsk, 2005).
J. Zhu, H. Zou, S. Rosset, and T. Hastie, “Multi-class AdaBoost,” Stat. Its Interface 2, 349–360 (2009).
Author information
Authors and Affiliations
Corresponding author
Additional information
This paper uses the materials of the report submitted at the 11th International Conference “Pattern Recognition and Image Analysis: New Information Technologies,” Samara, Russia, September 23–28, 2013.
Sergey M. Mikheev. Received his both master’s degree in integrated Navigation Systems engineering and PhD degree in System analysis, control, information processing from the Moscow Aviation Institute in 2008 and 2011 respectively. He began his career developing image processing algorithms for military multichannel optic systems. In 2012 he joined Samsung R&D Institute Russia, where he is responsible for tasks related to printing and scanning quality enhancement, imaging solutions for document management.
Ilya V. Kurilin. Received his MS degree in radio engineering from Novosibirsk State Technical University (NSTU), Russia in 1999 and his PhD degree in theoretical bases of informatics from NSTU in 2006. In 2007 Dr. I. Kurilin joined Image Enhancement Group, Signal Processing Lab, Samsung Research Center, Moscow, Russia where he is engaged in photo and document image enhancement projects.
Aleksey M. Vil’kin. Received his MS degree in mathematics from National Research Nuclear University MEPhI in 2011. From 2011 he is PhD student in MEPhI working on pattern recognition, page segmentation problems. In 2011 Aleksey joined Samsung Moscow Research Center, where he is engaged in on computer graphics, image and video processing projects.
Michael N. Rychagov. Graduated from the Moscow State University in 1986. Received his candidate’s degree in 1989 and doctorate degree in 2000. Scientific interests: image and video signal processing, biomedical visualization, ultrasonic measurement, engineering applications of neural networks. He is an associate professor at the Department of Theoretical and Experimental Physics (1998); professor at the Department of Biomedical Systems (2008); Director of the Algorithm Laboratory at Samsung Research Center in Moscow, Russia; and a member of IS&T and IEEE Societies.
Sang Ho Kim. Graduated from Seoul National University, Seoul, Korea, in 1990 and the Pohang University of Science and Technology, Pohang, Korea in 1992. Received his candidate’s degree in 2003. Scientific interests: electronic imaging systems, digital halftoning, image enhancement, color processing, and image compression.
Ho Keun Lee. Graduated from the Kyungpook National University in 2004. Scientific interests: image compression, image retrieval, and color image processing. A Senior Engineer at the Samsung Electronics Co. Ltd., Suwon, Korea.
Don Chul Choi. Graduated from the Sungkyunkwan University, Suwon, Korea in 1986. Scientific interests: SoC design, scanners, color processing, and image processing for electronic imaging systems. The vice president in IT Solutions Business Division.
Rights and permissions
About this article
Cite this article
Mikheev, S.M., Kurilin, I.V., Vil’kin, A.M. et al. Improving the print quality of screenshots. Pattern Recognit. Image Anal. 25, 674–684 (2015). https://doi.org/10.1134/S1054661815040173
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S1054661815040173