Fabric wrinkle evaluation using laser triangulation and neural network classifier

Wrinkling is one of the most important fabric performance properties, which are routinely evaluated in reference to a set of visual standards in the textile industry. The visual evaluation is unreliable and time-inefficient. An industrial need for objective and automatic evaluation methods has been...

Full description

Saved in:
Bibliographic Details
Published in:Optical Engineering 1999-10, Vol.38 (10), p.1688-1693
Main Authors: Su, Jie, Xu, Bugao
Format: Article
Language:English
Subjects:
fabric wrinkle evaluation
laser triangulation
neural-network classification
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Wrinkling is one of the most important fabric performance properties, which are routinely evaluated in reference to a set of visual standards in the textile industry. The visual evaluation is unreliable and time-inefficient. An industrial need for objective and automatic evaluation methods has been increasing markedly in the recent years. An image analysis system is developed to meet this need. The laser line triangulation method is used to measure the 3D surface data of a wrinkled fabric, and a neural network is built to execute the wrinkle classification with respect to the visual standard. Due to the directionality of wrinkles, a rotary stage is used to change the sample's orientation under the camera so that multiple images of the sample can be captured at different angles and more wrinkle data can be extracted for classification. The wrinkle classifications provided by the system are highly consistent with the visual standards, showing the potential for replacing human graders in fabric wrinkling evaluations. ©
ISSN:0091-3286
1560-2303
DOI:10.1117/1.602220