Automated inspection of solder joints on PC boards by supplementary processing of 3D and gray-level images

High-resolution, 3-D detection of the surface coordinates of printed circuit boards using optical 3-D sensors, combined with an available intensity image, offers a reliable way of judging the quality of wave-soldered components in through-hole and surface-mounting technologies. Global, 3-D pattern-recognition techniques, supported by gray-level operations that detect local disturbances in the brightness curve, produce a very comprehensive picture of the geometric shape and surface properties of solder joints. The capabilities of these methods are demonstrated for an automated solder-joint tester than can be integrated into the material and information flow of the production line. Results for recognition accuracy and processing speed are given.<
>