Computer Vision-Driven Fabric Categorization Using Deep CNNs for Knitted Textiles

Bangladesh's economy relies heavily on the readymade garments (RMG) industry, which serves as a central pillar of national economic activity, serving as a leading global exporter. This sector relied on labour-intensive and error-prone knitted fabric classification, which requires automation. Moreover, growing competition, shrinking delivery times, and the adoption of Industry 4.0 are accelerating the shift toward automation and intelligent fabric classification. This study employs a deep learning-based approach for classifying ten different types of knitted fabrics, which include Fleece, Nappy Jersey, Ottoman, Pique, Rib, Single Jersey, Slub, Terry, Velour towelling, and Pointelle Rib. A total of 2,104 industry-sourced real-world images of knitted fabrics were collected, which were then augmented to 20,493 grayscale images, each with a resolution of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML"…