Digital Color Images as a Tool for the Sustainable Use of Embroidery Elements from Folk Costumes

The aim of the research is to improve public’s assessment and understanding cultural values history Bulgaria. main issues related sustainable use elements heritage are defined, accessible literary sources digitization folklore reviewed. Shape indices, color, textural characteristics were obtained from digital color images Bulgarian folk costumes. most informative indices these features selected. A kernel variant principal component analysis (kPCA) method was used reduce data volume feature vector. Naïve Bayes classifier, discriminant analysis, support vector (SVM) for classification. classification accuracy assessed. In decorative costumes, it found that depended both on reducing separability classes data, depending classifier used. microscopic textile fabrics studied objects classifier. considered cases, a error below 10% using non-linear kPCA SVM with partition function. It proven results this development can be in creation modern cross-stitch patterns, clothing. practical application findings has potential benefit various stakeholders, including institutions, researchers, artisans, designers, general public, promoting deeper appreciation costume embroidery elements. Research continue direction preservation enriching future generations.