【作者】 谢佳豪；

【导师】 邱宗明；

【作者基本信息】 西安理工大学 ， 测试计量技术及仪器， 2004， 硕士

【摘要】 印刷电路板(PCB)是集成各种电子元器件的信息载体，在各个领域得到了广泛的应用。随着技术的不断发展和工艺水平的不断提高，PCB朝着层数更多、密度更高的方向发展。然而，如何在高度自动化的生产现场自动检测PCB板一直是研究人员的难题。本课题确定了利用光学摄像技术、微机控制技术、图像处理分析技术来自动检测PCB的方法，为PCB的自动检测提供了一个切实可行的方案。 本课题是在先前研究者的基础上进行的，他们的研究方向主要集中如何寻找图像处理算法来识别PCB的缺陷，从他们的研究成果来看，能检测出部分的印刷电路板的缺陷，如短路、断路、毛刺、缺损等等，但是采用图像处理技术检测出印刷电路板的缺陷，还只是印刷电路板自动检测系统的一个方面。除此以外，还有CCD高分辨率和大视场之间的矛盾，上下位微机并行系统如何控制印刷电路板图像自动采集，印刷电路板的线宽、线距和丢失线条等缺陷如何检测等问题还悬而未决，本课题将就如何结合计算机视觉技术、精密机械技术、自动控制技术和图像处理技术，如何解决图像采集高分辨率与大视场之间的矛盾，如何实现图像的自动拼接，如何实现两维工作台的精确定位，如何实现上下位机的准确通讯，如何检测线宽、线距缺陷和丢失线条等问题展开重点研究。实验证明，本课题较好的解决了上述几个问题，弥补了现行PCB检测系统的不足，为PCB的自动检测提供了一个较为完善的方案。更多还原

【Abstract】 Print Circuit Board (PCB) is an information carrier integrating various electronic components. It has been applied in different fields widely. With the continuous development of technology and improvement of the level of technics , PCB are prone to be more layers and more density . It is always a challenge to testing automatically high density PCB for researchers in the high automatic manufacture field. This paper has confirmed the project of inspecting PCB’ s defect automatically with optic photographic technique, computer control technique and image process technique, and has been proved to the a feasible method in the automatic testing of PCB.This paper was processing on the base of previous researcher , their research focus on how to seek for arithmetic of image process to inspect the defect of PCB , according to the result of their research , they pick out the part of defect from PCB , for example Short ,Cut , Burr, Mousebite and so on. But testing the defect adopting to image process is only a part of automatic inspection system for PCB. except that , many problem can’ t be solved , such as the conflict of CCD’ s high resolving power and big vision field , how to control the automatic gathering of PCB’ s image using master and slave computer parallel structure , how to inspect the defect of PCB such as width of circuit ,distance of circuit ,losing circuit and so on. The research aim at how to combine computer vision , precise machine , automatic control with image process , at how to resolve the contradiction between high resolving power of image gathering and wide vision field , at how to realize automatic mosaic of image , at how to realize precise orientation of two dimension worktable , at how to realize communication between master computer and slave computer , and at how to inspect the defect of line width , line distance and losing . later experimentation proved that the research solved above-mentioned problems successfully , compensates the shortcoming of the current testing system and makes it to be a complete PCB automatic testing system in function .更多还原