基于HALCON的PCB光学定位点的3种识别方法及比较

3 recognition methods and analysis of PCB Mark point based on HALCON

随着印刷电路板(PCB)的制造不断向着微型化方向发展,在其生产过程中,原有的人工检测形式逐渐被现代化的自动光学检测设备所代替。与此同时,在机器视觉代替人类视觉实现检测后,如何能快速识别并定位被检测PCB的光学定位点(Mark点)就成为整个检测过程中的一个重要问题,因为只有识别了光学定位点,才能进行后续的图像配准、匹配以及缺陷检测工作。HALCON作为一款功能强大的机器视觉软件,在工业图像处理方面有着不可替代的优势。笔者将HALCON软件应用到PCB的Mark点的识别过程中,创造出3种不同的定位算法,并对这3种算法的效率和应用范围进行了评价。

Along with the micromation trends of the Printed Circuit Board（PCB）,the original artificial inspection form is substituted by Automatic Optic Inspection（AOI） gradually.After the substitution,how to recognize the optic anchor point（Mark point） of the inspected PCB rapidly becomes a critical problem in the inspecting process,because it is the successful recognizing of the Mark point that ensures the following steps： image orientation,matching and defects inspecting.HALCON,as a powerful machine vision software,has its absolute merits in industrial image processing.The author creates 3 different arithmetics about the Mark point recognition,and makes some assessment from the perspectives of time-consumption and applicability.