Design and implementation of automatic optical surface defect inspection system of large aperture lenses

Surface quality of optical lenses is an important parameter to achieve optimal imaging quality. Surface imperfections such as scratches, pits, or digs which may be caused during manufacturing or handling induce light scattering and degrade the overall performance of the imaging system. For a lightweight lens, the pockets behind the lens will affect the defect judgment due to excellent optical transmittance. This article discusses the measurement flow of automatic optical inspection for large lenses, defect detection algorithm, and the practicality of application of this technique for inspection of large lenses. Surface defect detection module comprises of high dynamic range camera combined with telecentric imaging lens creating an imaging system with resolution of 2.7μm. Inspection of the whole surface before and after coating is carried out through concentric scanning and different illustrating angles. The images are captured with a 20% overlap between two consecutive images to ensure complete coverage of the whole surface. Image processing algorithms are applied to detect and classify any surface defects on the surface. Detection results consists of exact size, type, and location of the defects are analyzed by the system and summarized into an inspection report. Compared to classical manual inspection this system provides reproducible and objective detection results and allows users to perform revalidation after measurement is finished.