摘要
传统的光纤端面缺陷检测用的是人工检测方式,这种检测方式效率很低,检测结果的主观性很强。对光纤端面缺陷使用机器视觉检测,能极大地提高检测效率和检测准确性。首先将采集到的图像通过图像处理二值化,接着对纤芯中心进行定位,然后以纤芯中心为圆心对光纤端面进行不同的圆环检测区域划分。由于光纤端面上的缺陷有可能是暗色的或者是亮色的,因此为了区分二者,对每个区域检测时需要做不同的二值处理。如果有任意一个区域的检测不能通过,则这个光纤端面就是不符合要求的。结果显示,利用机器视觉进行光纤端面检测能够快速、精确地检测出缺陷所在位置及其大小。
The traditional defects inspection method of optical fiber end surface is manual inspection,which is a low efficiency method,and has subjective results.However,it is found that inspecting optical fiber end surface by machine vision can greatly improve the efficiency and accuracy.This article describes how to use machine vision for defects inspection in optical fiber end surface.Through processing the collected image,it is easy to get the binary image.After that,it needs to locate the centre of the core as the center of optical fiber end surface to divide the area.The defect of the optical fiber end surface may be dark or bright,in order to distinguish them,it needs to do different binary conversion.If any one area is failed,this optical fiber end surface is unqualified.The results show that using machine vision to inspect the optical fiber end surface can locate the defects and calculate the size of the defects rapidly and accurately.
出处
《现代电子技术》
2011年第19期136-139,143,共5页
Modern Electronics Technique
关键词
光纤端面
机器视觉
缺陷检测
图像处理
optical fiber end surface
machine vision
defects inspection
image processing
