A deep learning computer vision system trained on synthetic data is effective in detecting the following PCB defects:
During Reflow/SMT (Surface-mount technology), the system can detect tombstones, drawbridges, billboards, skew, twisting, reverse polarity, missing elements, and a lack of solder.
During wave soldering/THT (Through-hole technology), it can identify shorts, soldering bridges, insufficient wetting, and solder balling.
These defects and anomalies can be identified at various stages of the PCB assembly process, including screen printing, pick-and-place, solder paste application or printing, as well as before and after reflow oven or wave soldering stages. It's also applicable during existing manual or automatic optical inspection (MOI, AOI) or X-Ray Inspection (AXI).
This system can serve as an addition to or a retrofit for existing AOI (Automated Optical Inspection) systems. In doing so, it reduces false alarms while significantly increasing the defect detection rate.
