In the present work, the drape characteristics of second-hand textile fabrics were determined. The results can be used to implement the concepts of the circular economy. Automated software tools have been adapted and researched to apply the proposed methods and procedures for digital image analysis of used textile drape, which will be utilized to describe their shapes and predict their characteristics, as well as their classification into groups and assessment of classification accuracy in recognizing their elements. A radius-vector function was used to determine the main drape characteristics, such as the number of peaks, their size, and their location. Analytical models have been created for automated forecasting of the drape characteristics from used textiles, which can be applied to predict changes in these characteristics. It obtained an accuracy of 68–92% in the prediction of the main drape characteristics of used textiles. Due to changes in their main characteristics, the errors in classification and prediction increased by 10–15%. More complex computational procedures have been implemented to obtain a higher predictive power for second-hand textile fabrics. The results can be applied in the manufacture of new products such as curtains, tablecloths, napkins and fashion accessories.
Read full abstract