ABSTRACT
A novel text extraction method from graphical document images is presented in this paper. Graphical document images containing text and graphics components are considered as two-dimensional signals by which text and graphics have different morphological characteristics. The proposed algorithm relies upon a sparse representation framework with two appropriately chosen discriminative overcomplete dictionaries, each one gives sparse representation over one type of signal and non-sparse representation over the other. Separation of text and graphics components is obtained by promoting sparse representation of input images in these two dictionaries. Some heuristic rules are used for grouping text components into text strings in post-processing steps. The proposed method overcomes the problem of touching between text and graphics. Preliminary experiments show some promising results on different types of document.
- M. Aharon, M. Elad, and A. Bruckstein. K-SVD: an algorithm for designing overcomplete dictionaries for sparse representation. IEEE Trans. on Signal Processing, 54(11):4311--4322, 2006. Google ScholarDigital Library
- H. Barlow. Unsupervised learning. Neural Computation, 1(3):295--311, 1989. Google ScholarDigital Library
- E. J. Candès and D. L. Donoho. New tight frames of curvelets and optimal representations of objects with piecewise C 2 singularities. Comm. Pure Appl. Math., 57(2):219--266, 2002.Google ScholarCross Ref
- R. Cao and C. L. Tan. Text/graphics separation in maps. In D. Blostein and Y.-B. Kwon, editors, GREC, volume 2390 of LNCS, pages 167--177. Springer, 2001. Google ScholarDigital Library
- S. S. Chen, D. L. Donoho, and M. A. Saunders. Atomic decomposition by basis pursuit. SIAM Journal on Scientific Computing, 20(1):33--61, 1998. Google ScholarDigital Library
- G. Davis, S. Mallat, and M. Avellaneda. Adaptive greedy approximations. J. Constr. Approx., 13(1):57--98, 1997.Google ScholarCross Ref
- O. Deforges and D. Barba. A robust and multiscale document image segmentation for block line/text line structures extraction. In Proceedings of the 12th ICPR, volume 2, pages 306--310, 1994.Google Scholar
- D. L. Donoho. For most large underdetermined systems of linear equations the minimal l 1-norm solution is also the sparsest solution. Comm. Pure Appl. Math., 59(7):797--829, 2006.Google ScholarCross Ref
- D. L. Donoho and I. M. Johnstone. Ideal spatial adaptation by wavelet shrinkage. Biometrika, 81(3):425--455, 1994.Google ScholarCross Ref
- M. Elad, J. Starck, P. Querre, and D. Donoho. Simultaneous cartoon and texture image inpainting using morphological component analysis (MCA). Appl. Comput. Harmon. Anal., 19(3):340--358, 2005.Google ScholarCross Ref
- M. Fadili, J.-L. Starck, and F. Murtagh. Inpainting and zooming using sparse representations. Computer Journal, 52(1):64--79, 2009. Google ScholarDigital Library
- L. A. Fletcher and R. Kasturi. A robust algorithm for text string separation from mixed text/graphics images. IEEE Trans. Pattern Anal. Mach. Intell., 10(6):910--918, 1988. Google ScholarDigital Library
- H. Freeman and R. Shapira. Determining the minimum-area encasing rectangle for an arbitrary closed curve. Communications of the ACM, 18(7):409--413, 1975. Google ScholarDigital Library
- J. Gloger. Use of the Hough transform to separate merged text/graphics in forms. In Proceedings of the 11th ICPR, volume 1, pages 268--271, 1992.Google ScholarCross Ref
- R. C. Gonzalez and R. E. Woods. Digital Image Processing. Prentice Hall, 2nd edition, 2001. Google ScholarDigital Library
- T. V. Hoang, S. Tabbone, and N.-Y. Pham. Extraction of Nom text regions from stele images using area Voronoi diagram. In Proceedings of the 10th ICDAR, pages 921--925, 2009. Google ScholarDigital Library
- C. Jutten and J. Herault. Blind separation of sources, part 1: an adaptive algorithm based on neuromimetic architecture. Signal Process., 24(1):1--10, 1991. Google ScholarDigital Library
- C. P. Lai and R. Kasturi. Detection of dimension sets in engineering drawings. IEEE Trans. Pattern Anal. Mach. Intell., 16(8):848--855, 1994. Google ScholarDigital Library
- D. X. Le, G. R. Thoma, and H. Wechsler. Classification of binary document images into textual or nontextual data blocks using neural network models. Machine Vision and Applications, 8(5):289--304, 1995. Google ScholarDigital Library
- Z. Lu. Detection of text regions from digital engineering drawings. IEEE Trans. Pattern Anal. Mach. Intell., 20(4):431--439, 1998. Google ScholarDigital Library
- H. Luo and R. Kasturi. Improved directional morphological operations for separation of characters from maps/graphics. In K. Tombre and A. K. Chhabra, editors, GREC, volume 1389 of LNCS, pages 35--47. Springer, 1997. Google ScholarDigital Library
- B. A. Olshausen and D. J. Field. Emergence of simple-cell receptive field properties by learning a sparse code for natural images. Nature, 381:607--609, 1996.Google ScholarCross Ref
- B. A. Olshausen and D. J. Field. Sparse coding with an overcomplete basis set: a strategy employed by V1? Vision Research, 37(23):3311--3325, 1998.Google ScholarCross Ref
- W. Pan, T. Bui, and C. Suen. Text segmentation from complex background using sparse representations. In Proceedings of the 9th ICDAR, pages 412--416, 2007. Google ScholarDigital Library
- S. Sardy, A. G. Bruce, and P. Tseng. Block coordinate relaxation methods for nonparametric wavelet denoising. J. of Comput. and Graph. Stat., 9(2):361--379, 2000.Google Scholar
- M. Shensa. The discrete wavelet transform: wedding the à trous and Mallat algorithms. IEEE Trans. on Signal Processing, 40(10):2464--2482, 1992.Google ScholarDigital Library
- J. Song, F. Su, C.-L. Tai, and S. Cai. An object-oriented progressive-simplification-based vectorization system for engineering drawings: model, algorithm, and performance. IEEE Trans. Pattern Anal. Mach. Intell., 24(8):1048--1060, 2002. Google ScholarDigital Library
- J.-L. Starck, M. Elad, and D. L. Donoho. Image decomposition via the combination of sparse representations and a variational approach. IEEE Trans. on Image Processing, 14(10):1570--1582, 2005. Google ScholarDigital Library
- F. Su, T. Lu, R. Yang, S. Cai, and Y. Yang. A character segmentation method for engineering drawings based on holistic and contextual constraints. In Proceedings of the 8th GREC, pages 280--287, 2009.Google Scholar
- S. Tabbone, L. Wendling, and J.-P. Salmon. A new shape descriptor defined on the Radon transform. Comput. Vis. Image Underst., 102(1):42--51, 2006. Google ScholarDigital Library
- C. L. Tan and P. O. Ng. Text extraction using pyramid. Pattern Recognition, 31(1):63--72, 1998.Google ScholarCross Ref
- K. Tombre, S. Tabbone, L. Pélissier, B. Lamiroy, and P. Dosch. Text/graphics separation revisited. In D. P. Lopresti, J. Hu, and R. S. Kashi, editors, DAS, volume 2423 of LNCS, pages 200--211. Springer, 2002. Google ScholarDigital Library
- F. Wahl, K. Wong, and R. Casey. Block segmentation and text extraction in mixed text/image documents. Computer Graphics and Image Processing, 20(4):375--390, 1982.Google ScholarCross Ref
Index Terms
- Text extraction from graphical document images using sparse representation
Recommendations
Sparse representation with morphologic regularizations for single image super-resolution
Due to the fact that natural images are inherently sparse in some domains, sparse representation has led to interesting results in image acquiring, representing, and compressing high-dimensional signals. Based on the experiences and learned priors in ...
An antinoise sparse representation method for robust face recognition via joint l1 and l2 regularization
L1 or L2 regularization based representation is not antinoise enough.An antinoise sparse representation via joint L1 and L2 is proposed.The rationale of the objective function for fusion is analyzed.Recognition of noisy samples is evaluated as true ...
Image Denoising Using Low-Rank Dictionary and Sparse Representation
CIS '14: Proceedings of the 2014 Tenth International Conference on Computational Intelligence and SecurityIn this paper, we propose an image denoising model by using low-rank dictionary and sparse representation (LRSR). The K-SVD algorithm learns a universal dictionary for all patches in an image and the NLM exploits similarities of nonlocal patches, both ...
Comments