Automated PCB Defect Detection fabricated by CNC Machine: A Novel Approach Integrating Image Processing and Homography Transformation
Haroon Jan Khan, Department of Electrical and Computer Engineering, COMSATS University Islamabad, CUI, Abbottabad, Pakistan.
Jawad Saleem, Department of Electrical and Computer Engineering, COMSATS University Islamabad, CUI, Abbottabad, Pakistan.
Anooshay Niazi, Department of Electrical and Computer Engineering, COMSATS University Islamabad, CUI, Abbottabad, Pakistan.
Syed Nasir Shah, School of Energy and Electrical Engineering, Chang’an University, Xi’an 710064, Shaanxi, China.
Tariq, School of Energy and Electrical Engineering, Chang’an University, Xi’an 710064, Shaanxi, China.
Corresponding Author:
Haroon Jan Khan (hguruler@mu.edu.tr)
Abstract:
Achieving 100% defect-free Printed Circuit Boards (PCBs) is imperative in modern manufacturing. The demand for PCB inspection has surged, prompting advancements in CNC machine technology. This project focuses on comparing PCBs fabricated by a 3-axis CNC machine, employing image processing techniques, including Homography transformation, to detect defects. Misalignment poses a significant challenge during comparison. Utilizing a TB-6560 controller, PCBs are automatically fabricated, and images are captured for analysis in MATLAB. Various image processing algorithms, along with Homography transformation, are applied to detect defects and address misalignment issues. The study also evaluates the precision of the CNC machine.
Keywords:
Computer Numerical Control CNC; Printed Circuit Boards PCB’ XORing Images; Image Processing; Automated Inspection of PCB; MATLAB; Misalignment; Homography Transform; Hough Transform.