Analysis of the progress of machine vision-based surface defect detection technology

Abstract:
In industries such as semiconductor, PCB (Printed Circuit Board), automobile assembly, Liquid Crystal Display (LCD), 3C, photovoltaic cells, and textile, product appearance is inextricably linked to product performance. Surface defect detection is an important means to prevent defective products from entering the market. The use of machine vision for inspection is highly efficient and cost-effective and is the main direction of future development. This paper presents a statistical analysis of the number of defect detection objects, defect detection methods and defect detection literature over the last 30 years; finally, it points out the current problems with defect detection methods and provides an outlook for further work.