Browse > Article
http://dx.doi.org/10.6109/jkiice.2014.18.12.2831

Off-line Handwritten Flowchart Symbol Recognition Algorithm Robust to Variations Based the Normalized Dominant Slope Vector  

Lee, Gab-Seob (Department of Computer Information & Electronics, Gumi University)
Publication Information
Journal of the Korea Institute of Information and Communication Engineering / v.18, no.12, 2014 , pp. 2831-2838 More about this Journal
Abstract
This paper proposes the off-line handwritten flowchart symbol recognition algorithm by type and strength of a cross region of the straight line strokes that is extracted based the normalized dominant slope vectors. In the proposed algorithm, first of all, a connector symbol which consisted only curves is recognized by the special features, and the other symbols with straight line strokes are recognized by type and strength of a cross region, and that is extracted by extension of minimum bounding rectangle of the clusters of the normalized dominant slope vectors, and the straight line strokes of the symbols is extracted by the normalized dominant slope vectors. To confirm the validity of the proposed algorithm, the experiments are conducted for 10 different kinds of flowchart symbols that mainly used for computer program, and the number of symbols is 198. Experiment results were obtained the recognition rate of 99.5%, and the flowchart symbols is recognized correctly robust to variations, and then the proposed algorithm were found very effective for off-line handwritten flowchart symbol recognition.
Keywords
dominant slope vector; minimum bounding rectangle of cluster; cross region;
Citations & Related Records
연도 인용수 순위
  • Reference
1 M. Rusinol, L. Heras, J. Mas, O. R. Terrades, D. Karatzas, A. Dutta, G. Sanchez, and J. Lasods, CVC-UAB's participation in the Flowchart Recognition Task of CLEF-IP 2012. in CLEF 2012 evaluation labs and workshop [Online]. pp. 1-11, 2012.
2 M. Rusinol and L. Heras, O. Terrades, "Flowchart Recognition for Non-Texttual Information Retrieval in Patent Search," Information Retrieval, v. 17 no. 5/6, pp. 545-562, 2014.   DOI   ScienceOn
3 G. S. Lee, "An RDL Algorithm for the High Speed Extraction of Connected Components", Gumi College Rearch Bulletin, vol. 12, 2003.
4 K. Refaat, W. Helmy, A. Ali, M. AbdelGhany, and A. Atiya, "A New Approach for Context-Independent Handwritten Offline Diagram Recognition using Support Vector Machines," in 2008 International Joint Conference on Neural Networks, pp. 177-182, 2008.
5 Z. Yuan, H. Pan, and L. Zhang, "A Novel Pen-Based Flowchart Recognition System for Programming Teaching," Advances in Blended Learning, pp. 55-64, 2008.
6 M. C. Carlisle, T. A. Wilson, J. W. Humphries, and S. M. Hadfield, "RAPTOR : Introducing Programming to Non-Majors with Flowcharts," Journal of Computing Sciences in Colleges, vol .19 no. 4, pp. 52-60, April 2004.
7 X.H. Wu, "Automatic Conversion of Structured Flowcharts into Problem Analysis Diagram for Generation of Codes," Journal of Software, vol. 7, no. 5, pp. 1109-1120, 2012.
8 S. Ito, "Automatic Input of Flow Chart in Document Image," in 6th international conference on Software engineering, pp. 319-328, 1982.
9 B. G. Vasudevant, S. Dhanapanichkul, and R. Balakrishnan, "Flowchart Knowledge Extraction on Image Processing," Proceedings of International Joint Conference on Neural Networks, pp. 4075-4088, 2008.
10 W. Szwoch, "Recognition, Understanding and Aestheticization of Freehand Drawing Flowcharts," in ICDAR, vol. 2, pp. 1138-1142, 2007.