基于机器视觉的全自动探针台定位技术研究

发布时间：2018-04-01 14:41

本文选题：全自动探针台　切入点：机器视觉　出处：《浙江大学》2017年硕士论文

【摘要】：近年来,半导体制造工艺快速发展,微小芯片的市场需求急剧增加,芯片尺寸、引线间距和焊盘直径不断减小,因此对封装设备的运动精度要求越来越高,基于机器视觉的检测技术已成为IC装备精度保障的关键技术。本文研究了机器视觉在晶圆检测中的应用,基于视觉定位算法设计了全自动探针台中X/Y方向上的三个主要定位功能:晶圆圆心定位、晶圆位置自动校正以及针尖与针痕中心检测。全文主要包括以下内容:第一章阐述了机器视觉系统的应用及其研究现状,分析了机器视觉在晶圆检测中的应用情况,最后给出了论文的主要研究内容。第二章分析了全自动探针台视觉定位系统的三大主要功能需求,给出了全自动探针台视觉定位系统的总体设计方案,根据系统精度要求完成了视觉系统中相机、镜头的选型。第三章对比分析了四种经典边缘提取算法,在此基础上提出了多尺度结构边缘检测算法,实现了边缘点的准确定位;基于多尺度边缘检测算法设计了晶圆圆心检测方案,提高了晶圆圆心定位精度。第四章研究了 SURF、SIFT特征点检测算法,实验对比分析了两种算法的优劣性。根据实验结果,基于SIFT算法设计了晶圆位置校正方案,并通过实验验证了该方案的鲁棒性以及高精度。第五章分析了形态学图像处理算法以及多种轮廓中心求解算法,基于形态学图像处理算法以及中心矩算法设计了针尖与针痕中心检测方案,得出了针尖中心位置以及焊盘中心与针痕中心的位置差,保证了探针台自动对针功能的实现。第六章总结了全文主要研究内容,对后续研究工作进行展望。
[Abstract]:In recent years, with the rapid development of semiconductor manufacturing technology, the market demand for microchips has increased sharply, and the chip size, lead spacing and pad diameter have been reduced, so the motion accuracy of packaging equipment is required more and more. The detection technology based on machine vision has become the key technology of IC equipment precision guarantee. This paper studies the application of machine vision in wafer detection. Based on the visual localization algorithm, three main localization functions in the X / Y direction of the automatic probe table are designed: wafer center location, The main contents of this paper are as follows: in Chapter 1, the application of machine vision system and its research status are described, and the application of machine vision in wafer detection is analyzed. Finally, the main research contents of the thesis are given. In chapter 2, the three main functional requirements of the automatic probe platform vision positioning system are analyzed, and the overall design scheme of the automatic probe platform vision positioning system is given. According to the requirement of system precision, the selection of camera and lens in visual system is completed. In chapter 3, four classical edge detection algorithms are compared and analyzed. On this basis, a multi-scale structure edge detection algorithm is proposed to realize the accurate location of edge points. Based on the multi-scale edge detection algorithm, a wafer centroid detection scheme is designed, which improves the accuracy of the wafer centroid location. In Chapter 4, the detection algorithm of surfman sift feature points is studied, and the advantages and disadvantages of the two algorithms are compared and analyzed experimentally. According to the experimental results, the advantages and disadvantages of the two algorithms are compared and analyzed. Based on the SIFT algorithm, the wafer position correction scheme is designed, and the robustness and high precision of the scheme are verified by experiments. In chapter 5, the morphological image processing algorithm and various contour center solving algorithms are analyzed. Based on the morphological image processing algorithm and the center moment algorithm, the detection scheme of needle tip and needle mark center is designed, and the position difference between the tip center and the pin mark center is obtained. In chapter 6, the main contents of this paper are summarized, and the future research work is prospected.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;TN307

【参考文献】