BGA焊点缺陷在线自动识别方法研究

发布时间：2018-06-20 00:47

本文选题：缺陷识别 + X射线BGA焊点图像　；参考：《沈阳大学》2016年硕士论文

【摘要】：BGA封装是一种高集成度的现代集成电路器件封装方式,BGA封装器件的发展直接推动了芯片焊接技术的发展。但因为芯片内部焊点不可见并且焊点越来越小,焊接方式也越来越精密,造成了焊接缺陷增加的同时其检测技术也成为工业生产中的难题。然而在国内企业中,主要采用人工视觉检测的方法对BGA焊点缺陷进行检测,这大大增加了检测的不可靠度。因此如何提高BGA焊点缺陷检测的效率和准确性,以及提高检测效率和改善电子产品质量有着非常重要的研究价值和现实意义。本文重点完成以下工作:分析了BGA焊点的成像特点以及常用降噪方法对BGA焊点图像降噪时不佳现象的原因,并研究了PM模型在BGA焊点图像降噪的适用性;提出了以连通区域面积均值为阈值的新分割算法。主要研究了常用的阈值法并分析了它们分割不准确的原因,提出了以独立连通区域的面积均值为阈值然后用数学形态学平滑焊点边缘的新分割算法,不仅有效滤除线桥等干扰,而且使用此阈值结合经验来识别并滤除焊桥缺陷,统计出连通区域面积;对BGA焊点图像做直方图均衡化,分析了增强不明显的原因,提出了对每个焊点区域使用直方图均衡化单独增强,有效实现焊点区域与气泡区域具有明显的亮度区别;改进了梯度方向性滤波器。基于气泡区域对比度低、大小不同、位置随机等特点,设计了一种新的梯度方向性滤波器,成功滤除了非气泡区域;生成气泡区域的blob图从而提取出气泡部分,判断所有气泡的有效性,确定出有效气泡并统计其面积,然后计算出有效气泡占所在焊点的比值,再参照IPC标准结合实际生产要求对BGA焊点中的气泡缺陷进行缺陷识别。
[Abstract]:BGA packaging is a highly integrated and modern integrated circuit device packaging method. The development of BGA packaging device directly promotes the development of chip welding technology. However, the internal solder joint is invisible and the solder joint is becoming smaller and smaller, and the welding method is becoming more and more precise, which results in the increase of welding defects and its detection technology has become a difficult problem in industrial production. However, in domestic enterprises, manual visual inspection is mainly used to detect BGA solder joint defects, which greatly increases the reliability of detection. Therefore, how to improve the efficiency and accuracy of BGA solder joint defect detection, as well as improve the detection efficiency and improve the quality of electronic products has a very important research value and practical significance. The main work of this paper is as follows: the imaging characteristics of BGA solder joint and the causes of the poor noise reduction of BGA solder joint image by common noise reduction methods are analyzed, and the applicability of PM model in BGA solder joint image denoising is studied. A new segmentation algorithm based on the mean area of connected region is proposed. This paper mainly studies the commonly used threshold method and analyzes the reason why the segmentation is inaccurate. A new segmentation algorithm, which takes the area mean of the independent connected region as the threshold value and smoothes the solder joint edge with mathematical morphology, is proposed, which not only effectively filters the interference such as the wire bridge, but also the edge of the solder joint. Using this threshold and experience to identify and filter the defects of the welding bridge, the area of the connected area is calculated, and the BGA solder joint image is equalized by histogram, and the reason of the enhancement is analyzed. Using histogram equalization to each solder joint area is proposed to realize the obvious brightness difference between the solder joint region and the bubble area effectively and the gradient directional filter is improved. Based on the characteristics of low contrast, different size and random position of bubble region, a new gradient directional filter is designed to filter the non-bubble region successfully, and the bubble part is extracted by generating the blob diagram of bubble region. The validity of all bubbles is judged, the effective bubble is determined and its area is counted, then the ratio of the effective bubble to the solder joint is calculated, and then the bubble defect in BGA solder joint is identified according to the IPC standard combined with the actual production requirements.
【学位授予单位】：沈阳大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TN405

【相似文献】