The present paper proposes an artificial neural network model for the prediction of the degree of spirality of single jersey fabrics made from 100 % cotton conventional and modified ring spun yarns. The factors investigated were the yarn residual torque as the measured twist liveliness, yarn type, yarn linear density, fabric tightness factor, the number of feeders, rotational direction and gauge of the knitting machine and dyeing method. The artificial neural network model was compared with a multiple regression model, demonstrating that the neural network model produced superior results to predict the degree of fabric spirality after three washing and drying cycles. The relative importance of the investigated factors influencing the spirality of the fabric was also investigated.
In this paper, a computerized method has been proposed for automatic measurement and recognition of yarn wet snarls from an image of snarled yarn samples captured in a water bath. After image acquisition, image conversion and individual snarled sample extraction, the yarn profile function was extracted from the separated binary image. Fast Fourier Transform and Adaptive Orientated Orthogonal Projective Decomposition were then incorporated into a pattern recognition algorithm of yarn snarl features by treating the yarn profile function as a one-dimensional signal. In addition to the number of yarn snarl turns, the method was also accurate and efficient for the detection of yarn snarl height and width, which are unobtainable by the untwisting method. The effects of various factors on the yarn profile function were numerically examined, including distributions of yarn diameter and snarl, and the level of random noise.
Residual torque or twist liveliness of a single-spun yarn, among other factors, is the most prominent and fundamental factor contributing to the spirality of plain-knitted fabrics and the skewness of denim fabrics made from the yarn. In textile processing, yarn snarling caused by the twist liveliness is also considered a serious problem that leads to yarn breakage, the deterioration of yarn properties, and equipment malfunction [1]. Over recent decades, various testing apparatus and procedures have been developed for the measurement of twist liveliness. Generally, all of these methods are based on manual operation and can be classified into the three categories of direct, semi-direct, and indirect methods [2]. 1 Recently, computer vision has attracted the increasing attention of researchers in textiles and apparel. Attempts have been made to replace the conventional observation method with computer vision to resolve the limitation of human vision. The notable applications include pilling assessment [3][4][5][6], fabric texture analysis [7-10], fabric defects identification [11][12][13][14], and fiber morphological measurement [15,16]. In our earlier work [1], the methodology of a digital image-signal approach was proposed for the automatic measurement of yarn wet snarls from an image of snarled yarn samples captured in a water bath. After the image acquisition and pre-processing, the two digital signal-processing methods of Fast Fourier Transform (FFT) and Adaptive-orientated Orthogonal-projecAbstract This paper is the second part of a series reporting the recent development of a computerized method for the automatic measurement and recognition of yarn wet snarls from an image of snarled yarn samples captured in a water bath. In our earlier work, a digital image-signal approach for fully computerized yarn-snarl measurement was developed and the effects of various influencing factors on the recognition algorithms were numerically examined. In this paper, the feasibility and accuracy of the fully computerized method on the measurement of actual yarn wet snarls are evaluated through laboratory experiments. One hundred percent cotton ring spun single yarns of 7, 10, 16, and 20 Ne are prepared and used for the evaluation. In addition to the number of snarl turns per unit length, the snarl height and width of the yarn samples are also objectively measured by using the computerized method. The measurement results obtained by the computerized method are analyzed and compared with those measured manually by using a twist tester and an interactive computer method.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.