International Journal of Science, Engineering and TechnologyAuthors: Rofidatunnissa, Kusworo Adi, Catur Edi Widodo
Abstract: Woven fabrics are formed by the interlacing of warp (vertical) and weft (horizontal) yarns, with yarn density (threads per inch) being a critical parameter in textile quality assessment. This study aims to develop an automated yarn density measurement system using the Fast Fourier Transform (FFT), which analyzes periodic texture patterns in the frequency domain. The method involves frequency filtering and density estimation based on gray line profile intensity and was tested on 25 images for each weave type. Results show high accuracy for plain (0.96% warp error; 1.14% weft error) and twill (1.02% warp error; 1.57% weft error) weaves. However, satin weave exhibits a significant discrepancy between warp (31.98%) and weft (1.99%) errors, attributed to its unique structural characteristics—high yarn density, overlapping warp threads that obscure the weft, and a glossy surface that causes uneven light reflections, which affect image acquisition. While the method proves effective for most fabrics, accurately measuring warp density in satin remains a challenge. Nonetheless, the proposed approach has potential for industrial application to improve production efficiency in the textile industry.
DOI: http://doi.org/ijset.vol.13.issue3.198
