Rapid prototyping of optical lenses using the Isolation forest model for anomaly detection
Artykuł w czasopiśmie
MNiSW
100
Lista 2024
| Status: | |
| Autorzy: | Korga Sylwester |
| Dyscypliny: | |
| Aby zobaczyć szczegóły należy się zalogować. | |
| Rok wydania: | 2026 |
| Wersja dokumentu: | Elektroniczna |
| Język: | angielski |
| Numer czasopisma: | 5 |
| Wolumen/Tom: | 20 |
| Strony: | 1 - 17 |
| Impact Factor: | 1,3 |
| Web of Science® Times Cited: | 0 |
| Bazy: | Web of Science |
| Efekt badań statutowych | NIE |
| Materiał konferencyjny: | NIE |
| Publikacja OA: | TAK |
| Licencja: | |
| Sposób udostępnienia: | Otwarte czasopismo |
| Wersja tekstu: | Ostateczna wersja opublikowana |
| Czas opublikowania: | W momencie opublikowania |
| Data opublikowania w OA: | 1 kwietnia 2026 |
| Abstrakty: | angielski |
| The article presents the application of 3D printing technology in the production of optical lenses, which are used in fields such as mechanical engineering, computer science, and precision optics. The aim of the study was to design, manufacture, and verify the properties of transparent optical lenses made from photopolymer resins. The lenses were fabricated using stereolithography (SLA), a high-precision 3D printing method. The production process is described in detail, including the use of proprietary finishing tools that ensure precise lens processing crucial for achieving optimal optical properties. After printing, the lenses underwent finishing treatments including grinding, polishing, and UV curing, which significantly improved their optical quality. The best lenses demonstrated stable and high light transmission over a wavelength range from 380 to 1146 nm, with transmission values exceeding 90–92%. The optical properties of the lenses were assessed in a professional laboratory using a monochromator, allowing precise measurement and objective quality evaluation. A novel aspect of the study was the use of arti- ficial intelligence in quality control. The Isolation Forest model was employed to detect anomalies in the optical transmission data, identifying 79 anomalies under incandescent light and 81 under xenon light, approximately 5% of the dataset. To better interpret the detected anomalies, SHapley Additive exPlanations (SHAP) were used to understand which features most influenced anomaly classification. Combining advanced 3D printing technology with AI-driven anomaly detection represents an innovative and promising approach to optical lens manufacturing. This integration enhances defect detection, enabling more effective quality assessment and optimization of the production process. This method offers significant potential for rapid prototyping in precision optics applications. |
