Performance Scrutiny of Thinning Algorithms on Printed Gujarati Characters and Handwritten Numerals

  • Sanket B. Suthar
  • Rahul S. Goradia
  • Bijal N. Dalwadi
  • Sagar M. Patel
  • Sandip Patel
Conference paper
Part of the Lecture Notes in Networks and Systems book series (LNNS, volume 9)


We analyze the behavior of most common thinning algorithms on printed Gujarati text and handwritten numerals. We are focusing mostly on two types of algorithms: The first is serial thinning and second is parallel thinning. Thinning is more crucial when we focusing on structural feature-based character recognition. Thinned character reduced complication of the shape of the character. This analysis focuses on the actual result we get after applying serial and parallel thinning algorithms. Total five algorithms are used for experiments and applied on small, medium, big size of character data and on skewed character data. The results are useful where we designing classifiers for Gujarati text.


Gujarati text Handwritten numerals Thinning OCR Algorithms 


  1. 1.
    Hilditch CJ (1969) Linear skeletons from square cupboards. In: Meltzer B, Michie D (eds) Machine Intelligence, vol 4. Elsevier, New York: Amar, pp 403–420Google Scholar
  2. 2.
    Zhang TY, Suen CY (1984) A fast parallel algorithm for thinning digital patterns. Commun ACM 27(6):236–239CrossRefGoogle Scholar
  3. 3.
    Lu HE, Wang PSP (1985) An improved fast parallel algorithm for thinning digital patterns. Proceedings of the IEEE conference on computer vision and pattern recognition, pp 364–367Google Scholar
  4. 4.
    Stentifod FWM, Mortimer RG (1983) Some new heuristics for thinning binary handprinted characters for OCR. IEEE Trans Syst Man Cybern SMC-13(1)Google Scholar
  5. 5.
    Wang PSP, Hui L, Fleming Jr T (1986) Further improved fast parallel thinning algorithm for digital patterns. In: Wang PSP (ed) Computer vision, image processing and communication systems and applications, pp 37–40Google Scholar
  6. 6.
    Holt CM, Stewert A, Client M, Perrot RH (1987) An improved parallel thinning algorithm. CACM 30(2):156–160Google Scholar
  7. 7.
    Chin RT, Wan HK, Stover DL, Iverson RD (1987) A one pass thinning algorithm and its parallel implementation. Comput Vision, Graph Image Process 40(1):30–40Google Scholar
  8. 8.
    Pal S, Bhattacharyaa P (1989) A preserving one pass parallel thinning algorithm. Indian Institute of Management, Calcutta, Working paper series No 123(89)Google Scholar
  9. 9.
    Guo Z, Hall RW (1989) Parallel thinning with two subitration algorithm. CACM, 32(3):359–373Google Scholar
  10. 10.
    Chen CH, DeCurtins JL (1993) Word recognition in a segmentation-free approach to OCR. Proceedings of the second international conference on document analysis and recognition, pp 573–576. IEEE (1993)Google Scholar
  11. 11.
    Kardos Peter (2011) Sufficient conditions for order-independency in sequential thinning. Acta Cybernetica 20:87–100CrossRefMATHGoogle Scholar
  12. 12.
    Khalid S (2010) A universal algorithm for image skeletonization and a review of thinning techniques. Int J Apply Math Comput Sci 20(2):317–335MATHGoogle Scholar
  13. 13.
    Kalles D, Morries DT (1993) A novel fast and reliable thinning algorithm. Image Vis Comput 11(9):588–603CrossRefGoogle Scholar
  14. 14.
    Tarabek P (2008) Performance measurements of thinning algorithms. J Info Control Manag Syst 6(2)Google Scholar
  15. 15.
    Lam L, Lee SW, Suen CY (1992) Thinning methodologies-a comprehensive survey. IEEE Trans Pattern Anal Mach Intell 14(9) Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2018

Authors and Affiliations

  • Sanket B. Suthar
    • 1
  • Rahul S. Goradia
    • 2
  • Bijal N. Dalwadi
    • 3
  • Sagar M. Patel
    • 1
  • Sandip Patel
    • 1
  1. 1.Department of Information TechnologyCharotar University of Science and TechnologyChangaIndia
  2. 2.Department of Electronics and CommunicationG. H. Patel College of Engineering and TechnologyAnandIndia
  3. 3.Department of Information TechnologyBirla Vishvakarma MahavidyalayaAnandIndia

Personalised recommendations