• DocumentCode
    547290
  • Title

    Billet characters location in the complex scene

  • Author

    Hong, Hanyu ; Yang, Yijun ; Zhang, Xiuhua

  • Author_Institution
    Lab. for Image Process. & Intell. Control, Wuhan Inst. of Technol., Wuhan, China
  • Volume
    2
  • fYear
    2011
  • fDate
    10-12 June 2011
  • Firstpage
    753
  • Lastpage
    756
  • Abstract
    Locating the characters accurately is the primary and critical step of the billet characters recognition. To some ingredients of billet images in the production line, such as bad scene, complex illumination, high noise interference and unfixed position of billet characters in photos, which will lead to low in precision of billet characters location, a special algorithm complexed multilevel processing with the location of first rough and then accurate is presented in this paper. Noises are filtered through area filter, distance filter and projection filter gradually. To the special case that interference can´t filtered completely, locating the characters roughly and then multilevel processing are brought up. At last, billet characters will be located by means of projection function. Experimental results show that the proposed method can locate characters accurately from the complex background, and ensure the high recognition rate.
  • Keywords
    billets; character recognition; multiprocessing systems; production engineering computing; area filter; billet character location; billet character recognition; billet image; complex illumination; complex scene; distance filter; high noise interference; high recognition rate; multilevel processing; production line; projection filter; projection function; Billets; Character recognition; Filtering algorithms; Gravity; Information filters; Interference; character location; filter; first rough and then accurate; multilevel processing;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Science and Automation Engineering (CSAE), 2011 IEEE International Conference on
  • Conference_Location
    Shanghai
  • Print_ISBN
    978-1-4244-8727-1
  • Type

    conf

  • DOI
    10.1109/CSAE.2011.5952611
  • Filename
    5952611