Describing words by graphs
In this paper a new technique for handwritten word description is presented. It is based on the development of a description graph which results little sensitive to handwriting variability. The experimental results show the effectiveness of the description graph in retaining the basic information for the recognition of the word image.
- 1.K. Sakai, H. Asami, Y. Tanabe, “Advanced application systems for handwritten character recognition”, in From Pixels to Features III — Frontiers in Handwriting Recognition, S. Impedovo, J.C. Simon, (eds.), Elsevier Publ., 1992, pp. 175–186.Google Scholar
- 2.E. Lecolinet, O. Baret, “Cursive Word Recognition: Methods and Strategies”, in Fundamentals in Handwriting Recognition, S. Impedovo (ed.), NATO ASI Series, Springer Verlag 1994, pp. 235–263.Google Scholar
- 3.J.Hull, T.-K. Ho, J. Favata, V. Govindaraju, S. Srihari,”Combination of Segmentation-based and Wholistic Handwritten Word Recognition Algorithms”, in From Pixels to Features III — Frontiers in Handwriting Recognition, S. Impedovo, J.C. Simon, (eds.), Elsevier Publ., 1992, pp.229–240.Google Scholar
- 4.N.J. Naccache, R. Shingal, “SPTA:A proposed algorithm for thinning binary Patterns”, IEEE Trans. Syst. Man, Cybern., no. 3, 1984.Google Scholar
- 5.G. Dimauro, M.R. Grattagliano, S. Impedovo, G. Pirlo, “A System for Banckchecks Processing”, IEEE Computer Society Press, Vol. 4960-02, ICDAR '93, Tsukuba Science City-Japan, Oct. 1993, pp. 454–459.Google Scholar