Akademik Veri Yönetim Sistemi
3rd International Congress of Innovative Textiles Congress-ICONTEX 2022, Tekirdağ, Türkiye, 18 Mayıs 2022, ss.259-264
In parallel to the rapidly increasing population, fast fashion trends, the requirement for products with several characteristics increase the consumption of textile products and result in that natural fibers cannot fulfil the demand. Synthetic fibers have accommodated the natural fibers in the industry and have a growing market share. Filament yarns, made from synthetic fibers, are consisting of with being each and endless continuous filaments obtained by feeding molten polymers through the nozzles. In multifilament yarns, the number of filaments is a parameter that significantly affects the yarn and fabric properties. In the production of multifilament yarn, deviations in the filament numbers are observed especially due to the clogging of the nozzles during fiber spinning or fiber breaks caused by tension. In order to determine this, in addition to the standard physical tests such as unevenness, friction, breaking strength, and elongation after the yarn production, the number of filament determination test is also performed. In most cases, the numbers of filaments are manually counted by an operator, and no device or automation has been used yet. The aim of this study is to develop a novel method based on image processing for counting filament numbers. Within the scope of the recommended method, first of all, the filament yarns were charged with static electricity to ensure the separation of the fibers from each other. Then, the separated fibers were exposed to laser beam and each fiber and laser beam intersection were photographed. In the resulting photographs, each intersection point is assumed to represent single filaments and the total number of intersections represents the number of filaments in the yarn. The validation of the new method was provided by cross-sectional images of the filament yarns. With the proposed method, it is believed that errors caused by the operator will be reduced in the determination of the filament number and it will help automation in the future.