STM Journals Peer-reviewed research from Consortium e-Learning Network

Journal of Mechatronics and Automation

AUTOMATED PRINTED CIRCUIT BOARD (PCB) DEFECT DETECTION USING DIGITAL IMAGE PROCESSING TECHNIQUE

  1. M. Solomon
  2. S. E. Eke
  3. E. Ajulo
  4. M. Aibinu
  5. O. C. Ubadike

Abstract

This work presents a Digital Image Processing (DIP) solution involving the use of image filters and processing functions to analyse the reference and the test images of a Printed Circuit Board (PCB) to yield a resulting image in which open and short circuits have been highlighted. The problem solved was the dependence of other PCB defects detection methods on the absence of deformation in the test image and this was done using the “Bounding Rectangle Approach”. The method involved the use of bounding rectangles to identify those PCB trace regions of interest and narrow down the focus for defect detection to the actual size of the PCB. The overall sensitivity of fault detection by this system was made variable and could be set by the system operator. However, the sensitivity setting used in this work was 0.75 (75%). This approach greatly helped reduce computational cost by allowing usage of a low specification computer (Raspberry Pi) to achieve automation of the PCB defect detection process.

Keywords

Support