A numerical investigation of diffuse images for effective defect detection

• Published in: Proceedings of the Institution of Mechanical Engineers. Part B, Journal of engineering manufacture Vol. 216; no. 7; pp. 1073 - 1079
• Main Authors: Parker, J M, Hou, Z
• Format: Journal Article
• Language: English
• Published: London, England SAGE Publications 01.07.2002; Mechanical Engineering Publications; SAGE PUBLICATIONS, INC
• Subjects: Applied sciences; Artificial intelligence; Computer science; control theory; systems; Exact sciences and technology; Industrial metrology. Testing; Mechanical engineering. Machine design; Pattern recognition. Digital image processing. Computational geometry; defect detection; specular surfaces; reflectance model; image synthesis; Optical measurement; Computer vision; Image synthesis; Specular reflection; Artificial vision; Reflectance; Defect detection
• ISSN: 0954-4054; 2041-2975
• DOI: 10.1243/09544050260174274

Abstract Numerical simulation is a flexible, practical tool for efficiently investigating machine vision hardware/software configurations for myriad industrial applications and can significantly reduce the engineering time necessary to design and implement a system. In this paper, physically accurate image synthesis is used to investigate the ability of diffuse angle images to detect topographical defects reliably in smooth, highly specular coatings; a specular-plus-diffuse reflectance model provides the scientific underpinning for this hypothesis. The specific results presented in this study show that diffuse images of topographical defects yield high-contrast grey-scale images compared with diffuse images of similar coatings without topographical defects, facilitating effective defect detection and providing a foundation for characterizing the severity of defects present in a given region.