Exploring combined dark and bright field illumination to improve the detection of defects on specular surfaces
| Autor: | Paulo M.F. Forte, Eduardo J. Nunes-Pereira, Paulo Eduardo Reis Felgueiras, M.A. Sousa, Michael Belsley, Flávio P. Ferreira, Boris P. J. Bret |
|---|---|
| Přispěvatelé: | Universidade do Minho |
| Jazyk: | angličtina |
| Rok vydání: | 2017 |
| Předmět: |
Surface (mathematics)
Specular surfaces business.product_category Materials science media_common.quotation_subject 02 engineering and technology 01 natural sciences Bright-field illumination 010309 optics Optics 0103 physical sciences Contrast (vision) Köhler illumination Computer vision Specular reflection Electrical and Electronic Engineering Digital camera media_common Defect detection Science & Technology business.industry Mechanical Engineering Structured light Bright-field microscopy Dark-field illumination 021001 nanoscience & nanotechnology Atomic and Molecular Physics and Optics Electronic Optical and Magnetic Materials Critical illumination Automatic optical inspection Artificial intelligence 0210 nano-technology business |
| Zdroj: | Repositório Científico de Acesso Aberto de Portugal Repositório Científico de Acesso Aberto de Portugal (RCAAP) instacron:RCAAP |
| Popis: | An automatic optical inspection system for detecting local defects on specular surfaces is presented. The system uses an image display to produce a sequence of structured diffuse illumination patterns and a digital camera to acquire the corresponding sequence of images. An image enhancement algorithm, which measures the local intensity variations between bright- and dark-field illumination conditions, yields a final image in which the defects are revealed with a high contrast. Subsequently, an image segmentation algorithm, which compares statistically the enhanced image of the inspected surface with the corresponding image for a defect-free template, allows separating defects from non-defects with an adjusting decision threshold. The method can be applied to shiny surfaces of any material including metal, plastic and glass. The described method was tested on the plastic surface of a car dashboard system. We were able to detect not only scratches but also dust and fingerprints. In our experiment we observed a detection contrast increase from about 40%, when using an extended light source, to more than 90% when using a structured light source. The presented method is simple, robust and can be carried out with short cycle times, making it appropriate for applications in industrial environments. (C) 2016 Elsevier Ltd. All rights reserved. This research is sponsored by the Portugal Incentive System for Research and Technological Development. Project in co-promotion no 36265/2013 (Project HMIExcel - 2013-2015). info:eu-repo/semantics/publishedVersion |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|