3D computer-aided design systems have emerged as promising techniques for garment learning processing, virtual shopping, and fashion shows within the fashion and garment industries. However, the effective application of these digital systems requires precise characterization of fabrics, garment patterns, and human body shapes that accurately reflect the appearance and behavior of the garments in a digital environment. Creating a 3D digital garment involves inputting the corresponding digital fabric properties. Nevertheless, obtaining these measurements can be complex, often necessitating the involvement of well-trained technicians. In this study, our focus is on a simplified and automated technique for digitizing real fabrics. Specifically, we aim to find the most relevant digital fabric in the database of a 3D software by employing image processing and machine learning techniques to drape images.