Minority clothing pattern segmentation combined with intuitive fuzzy clustering optimization algorithm

Abstract

Ethnic costume patterns are important carriers of cultural heritage, containing rich historical and aesthetic values. Digitizing and preserving these patterns through segmentation techniques contributes to cultural education and resists the threat of cultural erosion brought about by globalization. To this end, this study introduces an intuitive fuzzy clustering optimization (IFCO) algorithm based on spatial neighborhood information (SNI) and membership constraint penalty term (MCPT) enhancement, which achieves robust segmentation of ethnic minority clothing patterns. By further integrating complementary spatial information (CSI) and membership connectivity mechanism (MDCM), a Minority Clothing Pattern Segmentation Combined with an Intuitive Fuzzy Clustering Optimization (MCPSC-IFCO) method is proposed. This method utilizes the Gaussian kernel distance metric (GKDM) to effectively suppress noise and capture complex patterns. The results showed that in mixed noise application environments of 5%, 8%, and 10%, the research method’s partition coefficient and partition entropy achieved excellent performance, with average values of 97.43% and 4.62%, respectively. In the segmentation environment of colored ethnic minority clothing patterns, the research method has greatly improved the segmentation performance compared to mainstream methods, with average segmentation coefficients and entropy of 97.04% and 5.29%, respectively. In the running efficiency results, the average running time was 36.84 ms. The above research results highlight the important significance of digital technology in protecting and promoting minority cultures, promoting cultural diversity and identity.