Article: Multi-layer elastic fabric cotton and linen fibre detection based on high precision computational vision Journal: International Journal of Materials and Product Technology (IJMPT) 2024 Vol.69 No.3/4 pp.244 - 264 Abstract: Multi-layer elastic fabrics have a complicated composition and interlacing that makes it challenging to detect them effectively using traditional testing techniques. In order to investigate the detection of cotton fabrics, this project can make use of high-precision optical tools, computer vision and image processing techniques, and cotton fabrics as the research object. In order to improve image quality and clarity, high-resolution camera images of textile surfaces are first collected, pre-processed using image processing technology, greyscale using a weighted average method, and noise removed using a median filter. After that, the technique is used to a novel wavelet transform approach that is based on wavelet transform, and it is used in real-world scenarios. To achieve the intelligent analysis and determination of the retrieved characteristic information, a recognition model of cotton and linen fibres is built based on this and the support vector machine (SVM) technique. Experimental results have shown that the accuracy, precision, and recall of the SVM-based model were 92.33%, 89.68%, and 94.77%, respectively, which can accurately detect cotton and linen fibres in multi-layer elastic fabrics. The research findings presented in this paper can offer fresh technical assistance for the textile industry's implementation of automated and intelligent manufacturing. Inderscience Publishers - linking academia, business and industry through research

Title: Multi-layer elastic fabric cotton and linen fibre detection based on high precision computational vision

Authors: Xiaofei Li

Addresses: School of Plastic Arts, Dankook University, Cheonan 31115, South Chungcheong Province, South Korea

Abstract: Multi-layer elastic fabrics have a complicated composition and interlacing that makes it challenging to detect them effectively using traditional testing techniques. In order to investigate the detection of cotton fabrics, this project can make use of high-precision optical tools, computer vision and image processing techniques, and cotton fabrics as the research object. In order to improve image quality and clarity, high-resolution camera images of textile surfaces are first collected, pre-processed using image processing technology, greyscale using a weighted average method, and noise removed using a median filter. After that, the technique is used to a novel wavelet transform approach that is based on wavelet transform, and it is used in real-world scenarios. To achieve the intelligent analysis and determination of the retrieved characteristic information, a recognition model of cotton and linen fibres is built based on this and the support vector machine (SVM) technique. Experimental results have shown that the accuracy, precision, and recall of the SVM-based model were 92.33%, 89.68%, and 94.77%, respectively, which can accurately detect cotton and linen fibres in multi-layer elastic fabrics. The research findings presented in this paper can offer fresh technical assistance for the textile industry's implementation of automated and intelligent manufacturing.

Keywords: computer vision; support vector machine; SVM; cotton and linen fibre detection; image pre-processing; local binary mode; greyscale treatment.

DOI: 10.1504/IJMPT.2024.145786

International Journal of Materials and Product Technology, 2024 Vol.69 No.3/4, pp.244 - 264

Received: 25 May 2024
Accepted: 07 Nov 2024
Published online: 23 Apr 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: Multi-layer elastic fabric cotton and linen fibre detection based on high precision computational vision

Keep up-to-date