A real-time machine vision system for solder paste inspection

As surface mount technology (SMT) component pitches decrease, quality inspection of the solder paste depositing is critical in the process of surface mounting. Currently, the inspection for solder pastes are manly performed by laser-based systems, however, it is not practical due to high cost and low inspection speed. In this paper, a real-time inspection approach based on the machine vision was presented. Firstly, a new fast image matching method was applied to align the PCB and determination of region of interest. Secondly, the images of the PCB are analyzed and the two-dimensional (2D) feature and a pseudo-3D feature are obtained. Finally, the artificial neuron network was established for classification the solder pastes. A prototype is developed and experiment results are presented, which illustrate the validity of the approach.