Elaboration of the hierarchical approach to segmentation of scanned documents images
The object of research is the process of recognizing the areas of scanned documents images. The paper proposes a hierarchical approach to the segmentation of scanned documents images. This approach is an image of a scanned document in the form of a multi-level structure. At each level of the structure, images containing structural regions are highlighted. Objects of the lower level strictly correlate with a certain area of the image of the upper level: areas of the photo and graphics correspond to the image containing the illustrations, and areas of text and background to the image containing both the text and the background at the same time. Using a hierarchical approach, it is possible to perform processing separately for each image area, namely: first, the areas of illustrations are highlighted on the original image of the scanned document using the analysis of connected components. Thus, the first level of the hierarchy consists of an image containing illustrations and an image containing text with a background. Then the areas of illustrations are divided into photos and graphics by splitting the areas of illustrations into blocks, and text areas are separated from the background using processing in the neighborhood of each pixel. Thus, the second level of the hierarchy is represented by images containing homogeneous areas: photos, graphics, text and background. The hierarchical approach to segmentation has reduced the processing time by an average of 80 times. The reduction in image processing time was due to the fact that at each level and in turn, in a separate part of the hierarchical structure, it was possible to take into account the structural features of a uniform image area corresponding to this level. And also choose the signs of identification of these areas with high computational efficiency, the use of which also reduced the processing time of the scanned document.
Shafait, F., Keysers, D., Breuel, T. M. (2008). Performance Evaluation and Benchmarking of Six-Page Segmentation Algorithms. IEEE Transactions on Pattern Analysis and Machine Intelligence, 30 (6), 941–954. doi: http://doi.org/10.1109/tpami.2007.70837
Kumar, S., Gupta, R., Khanna, N., Chaudhury, S., Joshi, S. D. (2007). Text Extraction and Document Image Segmentation Using Matched Wavelets and MRF Model. IEEE Transactions on Image Processing, 16 (8), 2117–2128. doi: http://doi.org/10.1109/tip.2007.900098
Acharyya, M., Kundu, M. K. (2001). Multiscale Segmentation of Document Images Using M-Band Wavelets. Lecture Notes in Computer Science, 510–517. doi: http://doi.org/10.1007/3-540-44692-3_62
Cesarini, F., Gori, M., Marinai, S., Soda, G. (1999). Structured document segmentation and representation by the modified X-Y tree. Proceedings of the Fifth International Conference on Document Analysis and Recognition. ICDAR’99 (Cat. No.PR00318), 563. doi: http://doi.org/10.1109/icdar.1999.791850
Baird, H. S., Moll, M. A., An, C., Casey, M. R. (2007). Document image content inventories. Document Recognition and Retrieval XIV. doi: http://doi.org/10.1117/12.705094
Vilkin, A., Egorova, M. (2010). Segmentatsiia otskanirovannykh dokumentov. GrafiKon'2010, 339–341.
Moiseev, N. N. (1981). Matematicheskie zadachi sistemnogo analiza. Moscow: Nauka, 487.
de Queiroz, R. L., Buckley, R. R., Xu, M. (1999). Mixed Raster Content (MRC) Model for Compound Image Compression. Visual Communications and Image Processing. San Jose, 3653, 1106–1117. doi: http://doi.org/10.1117/12.334618
Ishchenko, A., Polyakova, M., Kuvaieva, V., Nesteryuk, A. (2018). Elaboration of structural representation of regions of scanned document images for MRC model. Eastern-European Journal of Enterprise Technologies, 6 (2 (96)), 32–38. doi: http://doi.org/10.15587/1729-4061.2018.147671
Polyakova, M., Ishchenko, A., Huliaieva, N. (2018). Document image segmentation using averaging filtering and mathematical morphology. Telecommunications and Computer Engineering (TCSET). Lviv-Slavske. doi: http://doi.org/10.1109/tcset.2018.8336354
Polyakova, M., Ishchenko, A., Volkova, N., Pavlov, O. (2018). Combined method for scanned documents images segmentation using sequential extraction of regions. Eastern-European Journal of Enterprise Technologies, 5 (2 (95)), 6–15. doi: http://doi.org/10.15587/1729-4061.2018.142735
Magnier, B., Montesinos, P., Diep, D. (2011). Ridges and Valleys Detection in Images Using Difference of Rotating Half Smoothing Filters. Lecture Notes in Computer Science. Ghent, 261–272. doi: http://doi.org/10.1007/978-3-642-23687-7_24
Gusak, D. E., Ishhenko, A. V. (2019).Vydelenie tekstovykh fragmentov na izobrazhenii otskanirovannogo dokumenta. Suchasnі іnformatsіinі tekhnologіi. Odessa.
Sauvola, J., Kauniskangas, H. (1999) MediaTeam Document Database II, a CD-ROM collection of document images. University of Oulu.
GOST Style Citations
Copyright (c) 2019 Alesya Ishchenko, Vladyslav Zhuchkovskyi
This work is licensed under a Creative Commons Attribution 4.0 International License.
ISSN (print) 2664-9969, ISSN (on-line) 2706-5448