Abstract
Cellulosic yarns are the most fundamental and important materials for making a broad range of fashion structures and composites. Property of cellulosic yarns is mainly determined by their constitute fibers, and the internal structural properties, particularly the configurations and migration patterns of fibers inside the yarns. Tracer fiber technology is a popular method to measure fiber migration. The image mosaic and segmentation for two-viewed tracer fiber images are mainly conducted by manual operation. This paper is reporting the recent development of an intelligent method and automatic system for automatic mosaic and segmentation of tracer fiber images to analyze cellulosic yarn structural properties, including three-dimensional fiber configurations and migrations. Also a database composed of fifty series of tracer fiber images (total 872 images) with five different count densities of lyocell yarns (10Ne– 60Ne) was prepared and used to fully evaluate the qualities of the proposed image processing system with respect to conventional manual method. Evaluation results showed that the proposed method works well in automatic mosaic and segmentation for tracer fiber images for the intelligent structural analysis and evaluation. The proposed system presents a much higher efficiency than the conventional method, demonstrating a promising method and system for the structural analysis and evaluation of cellulosic yarns for fashion products. Copyright © 2021 The Author(s), under exclusive licence to Springer Nature B.V.
Original language | English |
---|---|
Pages (from-to) | 6739-6756 |
Journal | Cellulose |
Volume | 28 |
Issue number | 10 |
Early online date | 26 May 2021 |
DOIs | |
Publication status | Published - Jul 2021 |
Citation
Li, S. Y., & Fu, H. (2021). Image analysis and evaluation for internal structural properties of cellulosic yarn. Cellulose, 28(10), 6739-6756. doi: 10.1007/s10570-021-03900-zKeywords
- Cellulose fiber
- Image analysis
- Yarn properties
- Yarn structure
- Tracer fiber measurement