To obtain access to full text of journal and articles you must register!
- Article name
- HARDWARE-SOFTWARE COMPUTER VISION SYSTEM FOR RAPID SURFACE QUALITY INSPECTION OF INDUSTRIAL PRODUCTS
- Authors
- Garafutdinov A. A., , ayzatg@yandex.ru, FSBEI HE "Kazan National Research Technical University named after A. N. Tupolev - KAI", Kazan, Russia
- Keywords
- computer vision / quality control / surface defects / convolutional neural network / industrial inspection / image processing / hardware-software system
- Year
- 2026 Issue 3 Pages 41 - 45
- Code EDN
- SWRCJQ
- Code DOI
- 10.52190/2073-2597_2026_3_41
- Abstract
- The article presents a compact hardware-software computer vision system for rapid surface quality inspection of industrial products. The proposed system combines a camera, controlled LED illumination, a fixed holder, a computing unit, an operator interface, and a digital archive. The processing pipeline includes camera calibration, region-of-interest extraction, illumination normalization, noise suppression, calculation of statistical features, and classification by a convolutional neural network. A thresholding algorithm with morphological processing and contour analysis is used as a baseline. The model evaluation was conducted on a set of 520 images divided into training, validation, and test subsets. The convolutional model achieved an accuracy of 94.8 %, a defect recall of 92.3 %, and an average processing time of 0.18 s per image. It outperformed both manual visual inspection and the threshold-and-contour baseline.
- Text
- BUY for read the full text of article
- Buy
- 800.00 rub
