采用数字图像处理技术的条干均匀度在线检测系统  

The line inspection system for yarn evenness based on digital image processing techniques

作　　者：刘冀龙[1] 尹岗[1] 

机构地区：[1]内蒙古工业大学电力学院,内蒙古呼和浩特010051

出　　处：《微型机与应用》2016年第21期45-48,共4页Microcomputer & Its Applications

摘　　要：为了实现测量过程的自动化,提出了应用图像处理技术的条状物均匀度在线检测系统的设计方法。以交叉编译开发环境Qt Creator为核心,结合开源图像处理库OpenCV,实现了不规则条状物外尺寸的非接触式在线测量和实时显示。采用显微放大镜头,搭建了专用的光学系统;图像预处理采用最大类间方差法(Otsu),消除了不同光照强度对检测结果的影响;为了充分体现人机交互性而编写了友善的上位显示界面,采用MATLAB对采集到的数据进行曲线拟合,验证了分段直线拟合的可行性。实验结果表明,该装置结构简单,测量结果准确,使纺织品条干均匀度在线实时检测成为可能。In order to realize automatic measurement,a yarn inspection system based on digital image processing techniques was designed. Using Qt Creator as the core IDE,combined with the Open Source Computer Vision Library（ Open CV）,to realize the non-touching on-line measurement of yarn evenness and real-time transmission display. Based on the microscopy camera,the corresponding optical system was designed and constructed; in order to eliminate the effects of different light intensities,one adaptive threshold segmentation method was proposed,based on Otsu threshold method; to build a good Human-Computer Interaction experience,the friendly GUI of the system displaying was written and compiled. Finally the Sub-fitting results were carried out on the MATLAB. The experimental results show that the system has the advantages of simple structure,accurate measurement result and high precision,making the stem diameter line real-time inspection is possible.

关 键 词：数字图像处理 均匀度检测 最大类间方差法 分段直线拟合 

分 类 号：TH741[机械工程—光学工程]