Laser & Optoelectronics Progress, Volume. 53, Issue 4, 41501(2016)

Study of Shape Matching and Localization Algorithm Based on Polygon Fitting

Bai Yu*, Zhang Lichao, Shi Yusheng, and Yan Ran
Author Affiliations
  • [in Chinese]
  • show less

    In the control procedure of process based on machine vision, the key is the location of the special point on work piece. The difficulty lies in how to construct a shape descriptor that can meet the needs of shape matching, precision of location and real-time at the same time. To solve the problems above,a kind of shape matching and localization algorithm based on polygon fitting is proposed. The method is aimed at the planar two-dimension (2D) contour. The key points on the contour are gained by the polygon fitting processing, then the polygon is used to construct a shape description and the key points are used to make location, to obtain the transformation relationship between template contour and target contour. The results show that by using the proposed rotation invariant shape descriptor, high accuracy and good real-time performance of shape matching and location algorithm can be gained.

    Tools

    Get Citation

    Copy Citation Text

    Bai Yu, Zhang Lichao, Shi Yusheng, Yan Ran. Study of Shape Matching and Localization Algorithm Based on Polygon Fitting[J]. Laser & Optoelectronics Progress, 2016, 53(4): 41501

    Download Citation

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

    Category: Machine Vision

    Received: Oct. 12, 2015

    Accepted: --

    Published Online: Mar. 25, 2016

    The Author Email: Yu Bai (baiyupc@163.com)

    DOI:10.3788/lop53.041501

    Topics