Review of vision-based defect detection research and its perspectives for printed circuit board

Yongbing Zhou, Minghao Yuan, Jian Zhang*, Guofu Ding, Shengfeng Qin

*Corresponding author for this work

Research output: Contribution to journalReview articlepeer-review

14 Citations (Scopus)


The quality of the printed circuit board (PCB), an essential critical connection in contemporary electronic information goods, directly influences the efficiency and dependability of products. Therefore, any PCB defect should be identified promptly and precisely to avoid a product failure while it is in use. Numerous innovative methods based on machine vision, including automatic X-ray inspection (AXI), two-dimensional automated optical inspection (2D AOI), three-dimensional automated optical inspection (3D AOI), etc., are developed and used widely in PCB defect identification. Given the rapid research development in both image and vision computing and machine learning, two research questions are rising to us: (1) what is the current state-of-the-art in this research field? (2) what are the future research and development directions? To answer these two questions, this paper first systematically reviews the PCB visual detection methods and then explores the potential development trends. Firstly, we review and summarize the state of the art in research of the image data acquisition, image processing, feature extraction and feature recognition/classification methods for PCB defect detection, and then identify the commonly used method evaluation indicators and public data sets, and the execution feedback and optimization process from both visual inspection system and manufacturing system. Third, we identify the current challenges in PCB defect detection research and propose an intelligent PCB defect visual detection approach as a future potential development trend. Finally, how to implement the future potential development trend based on technology-driven and value-driven developments is discussed for intelligent manufacturing.

Original languageEnglish
Pages (from-to)557-578
Number of pages22
JournalJournal of Manufacturing Systems
Early online date3 Sept 2023
Publication statusPublished - 1 Oct 2023

Cite this