基于彩色图像处理与EDLines的数码印花缺陷检测系统  

Digital Printing Defect Detection System Based on Color Image Processing and EDLines

作　　者：黄乾玮 张团善[1] 周玲 汤锋 李乐乐 HUANG Qianwei;ZHANG Tuanshan;ZHOU Ling;TANG Feng;LI Lele(School of Mechanical and Electrical Engineering,Xi’an Polytechnic University,Xi’an 710600,China;Shaoxing Keqiao West-Tex Textile Industry Innovative Institute,Shaoxing,Zhejiang 312030,China)

机构地区：[1]西安工程大学机电工程学院,陕西西安710048 [2]绍兴市柯桥区西纺纺织产业创新研究院,浙江绍兴312030

出　　处：《轻工机械》2021年第4期74-79,共6页Light Industry Machinery

基　　金：国家自然科学基金(61701384);西安市现代智能纺织装备重点实验室资助项目(2019220614SYS021CG043)。

摘　　要：针对数码印花生产过程中由于喷头堵塞、电机偏差引起的PASS道缺陷问题,课题组设计了一套基于彩色图像处理与EDLines的数码印花缺陷检测系统。首先构建颜色补偿矩阵覆盖无关背景花案,增强缺陷与主色间差异性;然后分别在HSI颜色空间3通道采用自定义线型滤波锐化感兴趣区域并基于EDLines实现缺陷配准,并将3通道缺陷匹配结果进行区域融合和形态学处理;最后根据水平投影实现PASS道缺陷定位。实验结果表明:检测系统对印花织物表面缺陷的检测准确率达到96%以上,满足实际检测要求,为数码印花缺陷检测提供新的方法。A digital printing defect detection system based on color image processing and EDLines was designed to solve the problem of PASS defect caused by nozzle blockage and motor deviation in digital printing production. Firstly the color compensation matrix was built to overlay the irrelevant background pattern and enhance the differences between defects and the main color. Then the region of interest was sharpened in HSI color space three-channel by using custom linear filtering and the defect registration was realized based on EDLines. The three-channel defect registration results were regional integrated and morphologically processed. Finally, PASS channel defect location was realized according to the horizontal projection. The experimental results show that the detection accuracy of printed fabric surface defects can reach above 96%, which meets the actual detection requirements and provides a new method for the detection of digital printing defects.

关 键 词：缺陷检测 颜色空间 颜色补偿 EDLines 

分 类 号：TS107[轻工技术与工程—纺织工程] TH165[轻工技术与工程—纺织科学与工程]