Content area
Full Text
1.
Introduction
Electronic products are widely used in many industrial fields. As core components of these electronic products, an electronic component's function will directly affect the overall performance. Current electronic component assembly mainly uses soldering and surface mount technology on a circuit board; therefore, quality inspection of solder joints is very crucial (Wu et al. , 2009b; Lu and Zhang, 2008; Hongwei et al. , 2011). Solder joint defect types shown in Figure 1, such as insufficient solder, wrong component, component shifts, pseudo-solder, tombstoning, etc., that are found via traditional Manual Visual Inspection (MVI) of solder joint quality, rely on the detector. However, with the size of electronic components becoming ever smaller, their densities become higher, and it is difficult to make accurate judgments with the human eye. MVI relies on subjective experience, and the detection efficiency is often not high (Wu et al. , 2008, 2009a). After the implementation of lead-free production, the color of the surfaces of solder joints has become darker, so they are less suitable for manual inspection. Therefore, automated methods for detecting solder joint quality have become more and more essential (Hongwei et al. , 2011; Lu et al ., 2007).
Automatic optical inspection (AOI) systems use image acquisition to obtain the printed circuit board (PCB) image, and, through image feature extraction, processing and analysis, determine the quality of the solder joints. They are capable of testing high-density boards' and small electronic components' solder joints. Due to high detection standards, product quality, effective management and high efficiency, AOI is being used more and more for surface mount lines (Dom et al. , 1988; Yoda et al. , 1988; Wu et al. , 2009c; Hao et al. , 2013).
To research solder defect detection systems based on AOI of electronic components, the key issue is to solve the problems of image feature extraction of AOI, defect classification and algorithm running efficiency (Saenthon and Kaitwanidvilai, 2010; Jiang et al. , 2012; Takacs and Vajta, 2012; Wu et al. , 2013). To solve these problems, a series of studies have been conducted, mainly in the following areas:
* Image feature extraction and selection : The main features are the gray values (Kim and Cho, 1995; Kim et al. , 1999; Ong et al.