LED Die Cosmetic Inspection Using Automated Optical Inspection (AOI)
After LED die are created on the wafer, they must be inspected for surface defects, such as cracks, chips, and dark spots, as these imperfections can negatively affect the quality and performance of the LED. Because these types of defects vary and can occur in various locations, using rule-based machine vision is not feasible for high-speed inspection. Also, normal aberrations can occur that do not affect the quality of the LED die so it’s important for the system to ignore these minor defects. Given the size and volume of LED die that are processed daily, human inspection is neither efficient nor practical.
Cognex AI-based vision systems and software help manufacturers identify and classify real LED die defects. This advanced vision solution is trained using a series of images that represent both good and No Good (NG) results, so the software only flags significant defects. The locate tool identifies the Region of Interest (ROI). Once the ROI is defined, the defect detection tool identifies the defect within that area. The classification tool then categorizes defects. Using this information, production managers not only increase the yield of their finished LEDs, but also use the classification information to address and fix production issues, increasing profitability.