Laser & Optoelectronics Progress, Volume. 60, Issue 2, 0210007(2023)

Cigarette Package Image Registration Method Based on Feature of Single-Angle Region

Kaibin Wang, Ying Li*, Yinhao Li, and Zifen He
Author Affiliations
  • Faculty of Mechanical and Electrical Engineering, Kunming University of Science and Technology, Kunming 650500, Yunnan, China
  • show less
    Figures & Tables(14)
    SIFT registration. (a) Original image; (b) SIFT feature point matching; (c) registration result
    Image preprocessing. (a) Original image; (b) grayscale; (c) binary image
    Upper boundary point of cigarette package
    Upper edge of cigarette package. (a) Edge point; (b) edge line
    Result of rough image registration. (a) Rotation; (b) cutting
    Single⁃angle region
    Edge extraction results of Canny operator
    Results of boundary point screening and separation. (a) Boundary points; (b) points in C4; (c) boundary points
    Result of line fitting and corner position. (a) (b) line l3 and l4; (c) intersection of l3 and l4; (d) position of corner
    Cigarette package image acquisition equipment in operation
    Rough registration rendering
    Corner detection effect
    Original image and registration effect. (a) Template image; (b) real-time image; (c) SRR; (d) SIFT; (e) AKAZE; (f) ORB
    • Table 1. Registration speed

      View table

      Table 1. Registration speed

      AlgorithmRunning time /sSpeed /(frame·s-1
      SIFT(no rough registration)124.790.801
      SIFT98.591.014
      AKAZE44.172.26
      ORB9.3710.67
      SRR4.0524.69
    Tools

    Get Citation

    Copy Citation Text

    Kaibin Wang, Ying Li, Yinhao Li, Zifen He. Cigarette Package Image Registration Method Based on Feature of Single-Angle Region[J]. Laser & Optoelectronics Progress, 2023, 60(2): 0210007

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Image Processing

    Received: Sep. 7, 2021

    Accepted: Dec. 13, 2021

    Published Online: Dec. 9, 2022

    The Author Email: Ying Li (7114091@qq.com)

    DOI:10.3788/LOP212461

    Topics