HIGH-POWER LED ENCAPSULATION INSPECTION
Identify various defects during the LED encapsulation process
Graphical programming environment for deep learning-based industrial image analysis
High-power LEDs used for automotive lighting applications go through an encapsulation (potting) process after they are bonded to the substrate. This operation provides both protection for each die as well as a diffusing filter to soften the emitted light. High-speed dispensing machines fill each LED package with an encapsulant made from a mixture of epoxy and phosphor. Automated inspections are conducted post-dispensing to ensure consistent quality.
A variety of defects can come from this process, such as air bubbles, cracks, or excessive, insufficient, or missing encapsulant and foreign contaminants. Small defects within a certain tolerance are accepted; otherwise they are rejected and must be repaired or discarded. Given the variety and threshold levels of defects, this inspection process is too complex for traditional rule-based machine vision tools.
Cognex AI-based solutions help manufacturers of high-power LEDs identify and classify significant encapsulation defects. This advanced vision solution is trained using a series of images that represent both good and No Good (NG) results, enabling the software to filter out anomalies that are within an accepted range and flag only consequential defects. The locate tool identifies the Region of Interest (ROI) for inspection. Once the ROI is defined, the defect detection tool identifies any major flaws within that area.
The classification tool categorizes several types of defects. Using this information, production managers can increase the yield of their finished LEDs and use the classification information to address and fix production issues, which increases profitability and overall operational equipment efficiency (OEE).