Measuring Component Self-Alignment by Automatic Image Processing Method

Abstract

A method was developed for detecting the position of electronics components automatically before and after reflow soldering, allowing the analysis of component self-alignment. The method is based on image processing techniques like filtering, pattern recognition and particle swarm analysis. In the experiment, an FR4 based testboard was designed, which allows the measurement of component placement offsets by utilizing registration marks (fiducial points). Chip resistors (0603 size $-1.5\times 0.75 {mm})$ were placed onto the testboard with intentional component misplacement in × (0–800 μ m) and in y $({0-300} \mu m)$ direction. The position of the components of a test set was measured both before and after soldering, and the image processing method was customized based on the difference found between manual and automatic evaluation. Results showed that image processing techniques can successfully be applied to automatically evaluate component self-alignment. The error of the automatic position detection is only 10–15 μ m in average which means only a mistake of a 3–4 pixels during the image processing. The deviation of the error was also in the acceptable range taking also into account that results were validated against our former data created by manual evaluation of the exact positions.