RUTILEA Inc. (Headquarters: Kyoto City, Kyoto Prefecture) is a venture company that develops an open business model. Our product SDTest is an integrated solution of algorithms necessary for visual inspection, and we are proceeding with research and development with the goal of introducing it to more actual lines. We have developed a new imaging method that is optimal for visual inspection AI, and have achieved higher accuracy in AI visual inspection.
The newly developed AI-optimised imaging technology makes it possible to automate the visual inspection of workpieces that would be difficult to achieve by devising algorithms alone. From August of this year, we plan to sell a camera for visual inspection that incorporates this imaging algorithm.
AI optimised imaging
In recent years, efforts to automate visual inspection using AI have been attracting attention. However, AI technology alone may not be able to detect scratches or dents due to dirt or oil on the workpiece surface. To solve this problem, we developed an imaging method that is optimal for the algorithm we want to apply. As an example, after applying this imaging method to a workpiece that is difficult to visualise using normal imaging (Figure 1) or a workpiece with oil on the workpiece surface, Figure 2 shows the results of segmentation of dents. This method is also effective for multiple products such as die-cast products, resin products, electronic components, glass products, printed circuit boards, and cosmetics.
We would like to provide these as in-house developed hardware at low prices and contribute to better realization of production processes. If you interested, please feel free to contact us by email (email@example.com) or phone (075-746-2811) (Person in charge: Oshima). We would be glad to showcase this product and our achievements.
Figure 1 Visualisation of scratches on film-shaped workpieces that are difficult to image
Figure 2 Results of applying this method and segmentation algorithm to a bevel gear with oil stain