一种基于OpenCV的织物疵点检测方法  

A Method of Fabric Defects Detection Based on OpenCV

作　　者：陆趣趣[1] 陈传克[1] 徐晓龙[1] 虞佳[1] 贾小军[1] 

机构地区：[1]嘉兴学院数学与信息工程学院,浙江嘉兴314001

出　　处：《丝绸》2010年第4期38-41,共4页Journal of Silk

基　　金：浙江省教育厅大学生科技创新项目;嘉兴市科技局项目(2008AY2014)

摘　　要：在织物疵点自动检测开发中,传统的图像处理代码编写繁琐、效率不高。OpenCV具有较强的图像处理能力且提供了丰富的图像处理函数,可以把OpenCV运用到织物疵点检测上。以断经、纬疵样本为例,提出在OpenCV环境下采用阈值分割提取疵点图像,利用形态学技术实现噪点分离及断线连接。边缘检测实现了疵点图像在原图像中的准确定位。实验结果表明,OpenCV有简化代码、提高编程效率的图像处理强大功能,疵点检测结果准确,效率高。In the fabric defects automatic detection development, traditional image processing code compilation is tedious and inefficient. OpenCV has powerful image processing ability and providing rich image processing functions, which may be utilized to process fabric defects detection. This paper took broken end, irregular pick fabric as the examples. Under OpenCV environment, using the thresholding segmentation to extract fabric defects image. Morphological techniques can remove noise pixels and connect broken lines. Edge detection achieves positioning defect image in the original image. The experimental results showed that OpenCV had the powerful functions of simplifying code, raising efficiency to process images, and the results of fabric detection were accurate and efficient.

关 键 词：OPENCV 疵点检测 阈值分割 

分 类 号：TS101.9[轻工技术与工程—纺织工程] TN911.73[轻工技术与工程—纺织科学与工程]