Inspection that is uncertain about the judgment of scratches and dents | Gemini eye

Learn the range of good products and reduce the variability in judgment by the viewer.

In visual inspections of automotive parts and mass-produced components, there are moments of uncertainty about how much damage or dents can be tolerated. When the judgment varies depending on the inspector, it increases the burden of verification in subsequent processes and quality control. "Gemini eye" is an appearance inspection AI that learns from good products and detects areas that are "different from usual." It can be evaluated with a small amount of good product data, eliminating the need to collect a large number of defective products. It addresses issues such as dents on nuts, rust on bearings, and cuts on rubber parts. Abnormal areas are displayed on a heat map, allowing on-site verification of where defects were identified. The judgment results can be confirmed through images, numbers, and graphs, making it effective for sharing inspection standards and training. In addition to detecting unknown defects, it also supports the display of known defects by type. Consultation is available from the selection of imaging equipment to inspection devices, allowing for a phased approach from assisting existing visual inspections to automation. Proposals are made with an eye toward operations tailored to line conditions and target objects. 【Features】 ■ AI learning possible with only good products ■ Detection of scratches, dents, rust, etc. ■ Abnormal areas displayed on a heat map ■ Proposals for imaging equipment and inspection devices *For more details, please refer to the PDF document or feel free to contact us.

Image Processing Software