FDAS: A Knowledge-based Framework for Analysis of Defects in Woven Textile Structures |
| |
Authors: | K. Srinivasan P. H. Dastoor P. Radhakrishnaiah Sundaresan Jayaraman |
| |
Affiliation: | School of Textile and Fiber Engineering, Georgia Institute of Technology , Atlanta, GA, U.S.A. |
| |
Abstract: | Defects in woven textile structures have been analyzed, and a novel scheme for their classification based on their visual attributes is proposed. The proposed scheme can serve as the underlying framework for a vision-based inspection system. The classification framework has been incorporated in software. The resulting knowledge-based system (FDAS — Fabric Defects Analysis System) identities defects, assigns probable causes for the defects, and suggests plausible remedies to avoid them. The system has been tested with actual fabric defects and has performed well. In addition to being used on the shopfloor, FDAS can be used for training new operators in fabric inspection in weaving and apparel-manufacturing plants. |
| |
Keywords: | |
|
|