Browse > Article
http://dx.doi.org/10.3745/KIPSTB.2011.18B.4.221

Comic Image Normalization using the gradient Radon Transform based on OpenCL implementation  

Kim, Dong-Keun (공주대학교 컴퓨터공학부)
Jeon, Hyeok-June (한국항공우주연구원)
Hwang, Chi-Jung (충남대학교 컴퓨터공학과)
Publication Information
The KIPS Transactions:PartB / v.18B, no.4, 2011 , pp. 221-230 More about this Journal
Abstract
Digital comic images are one of popular contents on the Internet. Usually, they are scanned from comic books by digital scanners. Without post-processing, they may have different sizes, skews and margins other than contents at the boundary. To normalize the size of their contents without the skews and margins is an important step in comic image analysis and application such as content-based comic image retrieval system. In this paper, we propose a method to detect a box frame in comic images by extracting of line segments using the gradient Radon transform. The box frame in comic images is the maximum rectangle which consists of contents without margins. We use the detected box frame to normalize the size of comic images and to make them no skew. In addition, the proposed method is implemented by OpenCL to speed up the detection of the line segments. Experimental results show that our proposed method effectively detects the box frame in comic images.
Keywords
Radon Transform; OpenCL; Comic Image;
Citations & Related Records
Times Cited By KSCI : 2  (Citation Analysis)
연도 인용수 순위
1 Duda, R. O. and P. E. Hart, "Use of the Hough Transformation to detect Lines and Curves in Pictures," Communications of the ACM, Vol.15, pp.11-15, 1972.   DOI
2 D. Ballard and C. Brown Computer Vision, Prentice-Hall, 1982.
3 William H. Press, "Discrete Radon transform has an exact, fast inverse and generalizes to operations other than sums along lines," Proceedings of the National Academy of Sciences(PNAS), Vol.103, No.51, pp.19249-19254, 2006.   DOI   ScienceOn
4 Peyrin.F, Goutte.R, "Image invariant via the Radon Transform" Image Processing and its Applications, pp.458-461, 1992.
5 Yamada et al., "Mamoru Endoo, and Shinya Miyazaki. Comic image decomposition for reading comics on cellular phones," IEICE transactions on information and systems, E87-D(6), pp.1370-1376, 2004.
6 OpenCV Reference Manual, http://sourceforge.net/projects/opencvlibrary/
7 L. Xu, E. Oja, and P. Kultanan, "A new curve detection method: Randomized Hough transform (RHT)", Pattern Recognition Letter, pp.331-338, 1990.
8 E. Montseny et al, "Edge Orientation-based Fuzzy Hough Transform (EOFHT)," European Society for Fuzzy Logic and Technology(EUSFLAT) Conference Proceedings, 2003.
9 김동근, OpenCV programming, 가메출판사, 2010.
10 Intel OpenCL SDK, http://software.intel.com/en-us/articles/intel-opencl-sdk/
11 http://www.khronos.org/opencl/
12 N. Kiryati, Y. Eldar, A. M. Bruckstein, "A probabilistic Hough transform", Pattern Recognition, Vol.24, No.4, pp.303-316, 1991.   DOI   ScienceOn
13 T. S. Durrani and D. Bisset, "The Radon transform and its properties", GEOPHYSICS, Vol.49, No.8, pp.1180-1187, 1984.   DOI   ScienceOn
14 A.S.M Shihavuddin et. al, "Road Boundary Detection by a Remote Vehicle Using Radon Transform for Path Map Generation of an Unknown Area", International Journal of Computer Science and Network Security(IJCSNS), Vol.8 No.8, pp.64-69, 2008.
15 Si-Yu Guo et al, "Probabilistic Hough transform for line detection utilizing surround suppression," International Conference on Machine Learning and Cybernetics, pp.2993-2998, 2008.
16 김동근, 양승범, 황치정,"그래디언트 라돈변환을 이용한 만화영상의 외곽 경계사각형 검출," 멀티미디어 학회 논문지, 제 14권 4호, pp.538-545, 2011.
17 Kenji et al., Juichi Miyamichi, "Layout Analysis of Tree-Structured Scene Frames in Comic Images," International Joint Conference on Artificial Intelligence, pp.2885-2890, 2007.
18 한은정, 장창혁, 정기철, " 다층 신경망을 이용한 모바일 자동변환 시스템," 멀티미디어 학회 논문지, 제 12권 2호, pp.272-280, 2008.   과학기술학회마을
19 Yasuto Ishitani, "Document layout analysis by interaction between data-driven processing and concept-driven processing", Journal of Information Processing Society of Japan, Vol.42, No.11, pp.2711-2723, 2001.