Acta Optica Sinica, Volume. 41, Issue 5, 0515003(2021)

Edge Detection and Repair of PCBA Components Based on Adaptive Canny Operator

He Yan*, Qifeng Zhao**, Min Xie, and Xiaoling Li
Author Affiliations
  • Liangjiang Artificial Intelligence Academy, Chongqing University of Technology, Chongqing 401147, China
  • show less

    The integrity of edge detection of printed circuit board assembly (PCBA) components directly affects the visual measurement accuracy of the size and gap of components obtained by assembly robots. Aiming at the problem that the edges of objects in dense areas extracted by the Canny operator have obvious edge adhesion and missing phenomena, a new method of PCBA component edge detection with high integrity is proposed. First, for the edge pixels determined by the original Canny operator based on the gradient of the pixel gray value, an adaptive threshold non-maximum suppression method with a 3×3 neighborhood window is proposed to effectively avoid the adhesion of adjacent edges in dense areas. Second, based on the gradient field of the Snake curve after parameterization by the EPGVF Snake model, combined with the edge fidelity term, a neighborhood window adaptive judgment method for determining whether the edge is broken is proposed. Finally, a pixel replacement template is used to fill the broken edge pixels to effectively retain the weak edges and avoid breaks. Experimental results show that the proposed method can avoid the adjacent edge adhesion and weak edge missing effectively in the dense area of PCBA, and the integrity of the preserved edge is improved about 15.5% compared to other methods.

    Tools

    Get Citation

    Copy Citation Text

    He Yan, Qifeng Zhao, Min Xie, Xiaoling Li. Edge Detection and Repair of PCBA Components Based on Adaptive Canny Operator[J]. Acta Optica Sinica, 2021, 41(5): 0515003

    Download Citation

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

    Category: Machine Vision

    Received: Sep. 7, 2020

    Accepted: Nov. 2, 2020

    Published Online: Apr. 7, 2021

    The Author Email: Yan He (yanhe@cqut.edu.cn), Zhao Qifeng (zqf@2018.cqut.edu.cn)

    DOI:10.3788/AOS202141.0515003

    Topics