Automatic Detection of Woven Fabric Defects by One-class Support Vector Method

摘要：

A support vector method called support vector data description for one-class classification is introduced into the research area of woven fabric defects detection. To validate the feasibility of the proposed method, detection experiments have been made on defective fabric samples with various texture backgrounds. Expected false alarm rates are obtained, and the missing rate as well as the false alarm rate can be kept at low levels simultaneously.

关键词： woven fabric defects detection support vector method one-class classification

作者: BU Hong-gang WANG Jun HUANG Xiu-bao

作者单位: College of Textiles, Donghua University, China Key Laboratory of Textile Science & Technology of Ministry of Education, China

会议类型: 国际会议

会议名称: International Conference on Fibrous Materials 2009,the Fiber Society 2009 Spring Conference(纤维材料国际研讨会暨纤维协会2009年春季会议)

会议地点: 上海

会议语种:英文

页码: 672-675

在线出版日期: 2009-05-27（万方平台首次上网日期，不代表论文的发表时间）