Intelligent Positioning Algorithm of Solder Joints Based on Color Features

Xie Hong-wei,Zhang Kun,Kuang Yong-cong,Zhang Xian-min,Pei Hai-long
DOI: https://doi.org/10.3969/j.issn.1000-565X.2013.03.017
2013-01-01
Abstract:In order to improve the automation degree of solder joint positioning by using the existing automatic optical inspection systems,an intelligent positioning algorithm based on color features is proposed.In this algorithm,the initial color thresholds of solar joint segmentation are calculated by statistically analyzing the gray histograms of solder joint regions,and a minimum pixel loss method is used to correct the initial thresholds and to further obtain the optimal threshold.Thus,the solder joint color is automatically extracted.Moreover,during the positioning,the ima-ges of solder joints are binarized with the obtained threshold,and the optical location of solder joints are fast searched by using the optimized maximum area method.Experimental results show that the proposed algorithm is effective because it helps to obtain properly-extracted solder joint images with small positioning error and high positio-ning speed that meets the practical application requirements well.
What problem does this paper attempt to address?