Liu Jingyuan, Fang Liyong, Hu Dongcai, Qi Xiaoshi, Deng Yangquan
{"title":"A new method to extract BGA solder balls in complex background X-ray image based on coordinate transformation","authors":"Liu Jingyuan, Fang Liyong, Hu Dongcai, Qi Xiaoshi, Deng Yangquan","doi":"10.1109/ITNEC.2019.8729540","DOIUrl":null,"url":null,"abstract":"In the detection of quality of BGA soldering based on AXI (Automated X-Ray Inspection), it’s critical important to extract solder balls accurately. Although the existing methods can extract most of the solder balls in X-ray images, they are sensitive to parameters and could only be used in the case of simple background. In order to solve this problem, a new method based on coordinate transformation is proposed in this paper. In this method, original X-ray images in Cartesian coordinate system are transformed in to polar coordinate firstly, so the edge of solder ball would be transformed from a curve to a straight line. Secondly a nonparametric edge detection could be used to extract the edge of solder balls. Result in Polar coordinate system would be transformed into original Cartesian coordinate system, finally. Through the comparison of two experiments, it is confirmed that solder balls in X-ray image could be extracted by proposed method precisely, while the proposed method has much more adaptability than other traditional methods. When proposed methods combined with other method such as the detection of voids in solder balls, the detection would be effective and improve the performance of AXI system ultimately.","PeriodicalId":202966,"journal":{"name":"2019 IEEE 3rd Information Technology, Networking, Electronic and Automation Control Conference (ITNEC)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE 3rd Information Technology, Networking, Electronic and Automation Control Conference (ITNEC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ITNEC.2019.8729540","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
In the detection of quality of BGA soldering based on AXI (Automated X-Ray Inspection), it’s critical important to extract solder balls accurately. Although the existing methods can extract most of the solder balls in X-ray images, they are sensitive to parameters and could only be used in the case of simple background. In order to solve this problem, a new method based on coordinate transformation is proposed in this paper. In this method, original X-ray images in Cartesian coordinate system are transformed in to polar coordinate firstly, so the edge of solder ball would be transformed from a curve to a straight line. Secondly a nonparametric edge detection could be used to extract the edge of solder balls. Result in Polar coordinate system would be transformed into original Cartesian coordinate system, finally. Through the comparison of two experiments, it is confirmed that solder balls in X-ray image could be extracted by proposed method precisely, while the proposed method has much more adaptability than other traditional methods. When proposed methods combined with other method such as the detection of voids in solder balls, the detection would be effective and improve the performance of AXI system ultimately.