数字图像处理技术在织物瑕疵检测中的应用研究  

Research on the Application of Digital Image Processing Technology in Fabric Defect Detection

作　　者：杨骏浩 傅丹妮 陈盈娜 YANG Junhao;FU Danni;CHEN Yingna

机构地区：[1]宁波大学科学技术学院

出　　处：《丝网印刷》2023年第15期23-25,共3页Screen Printing

摘　　要：纺织业是我国传统行业的支柱性产业之一,在我国的经济构成与发展中起着重要作用。然而,织物瑕疵点的存在影响产品品控,是纺织业面临的主要难题之一。文章结合数字图像处理技术,围绕破损型织物瑕疵,通过图像预处理、特征提取和孤立森林异常检测实现了织物瑕疵点的有效检测,为织物瑕疵点的快速与稳定检测提供了完整的处理流程,有助于提升企业产品品控,降低成本,提升生产质量。The textile industry is one of the pillar industries in China’s traditional industries,and plays an important role in the country’s economic composition and development.However,the presence of fabric defects affects product quality control and is one of the main challenges facing the textile industry.This article combines digital image processing technology,focuses on the broken fabric defects,and realizes the effective detection of fabric defects through image preprocessing,feature extraction,and isolated forest anomaly detection.It provides a complete processing fl ow for the rapid and stable detection of fabric defects,which helps to improve product quality control,reduce costs,and enhance production quality.

关 键 词：织物疵点检测 数字图像处理 孤立森林 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TS107[自动化与计算机技术—计算机科学与技术]