纺织高校基础科学学报2025，Vol.38Issue(5)：88-97,10.DOI:10.13338/j.issn.1006-8341.2025.05.010
基于特征选择与聚类优化的羊绒羊毛分类方法
A classification method for cashmere and wool based on feature selection and clustering optimization
摘要
Abstract
To address the challenge of low inter-class variance and high intra-class variance in cashmere and wool fiber image classification,this paper proposed a fine-grained classification method that integrates feature selection and clustering optimization.First,morphological,textur-al,and keypoint features of cashmere and wool fibers were extracted to capture subtle differences among fiber types.Then,a feature selection method based on intra-class and inter-class distance was employed to identify highly discriminative features and reduce redundant information.Final-ly,an improved K-means clustering algorithm was introduced,incorporating Mahalanobis distance for intra-class similarity and Euclidean distance for inter-class separation to identify and mitigate the influence of outliers.This enhances intra-cluster compactness and inter-cluster separability.The experimental results demonstrate that the proposed method achieves a classification accuracy of 98.92％on the cashmere/wool fiber image dataset,representing a 3.04％improvement over the baseline model.Simultaneously,the intra-class dispersion decreases by approximately 37％o-verall,while the inter-class separation increases by about 26％,indicating that the proposed meth-od exhibits significant advantages in fine-grained fiber classification.关键词
细粒度纤维图像分类/类内马氏距离/类间欧式距离/特征选择/改进K-meansKey words
fine-grained fiber image classification/intra-class mahalanobis distance/inter-class euclidean distance/feature selection/improved K-means分类
信息技术与安全科学引用本文复制引用
顾梅花,候嘉乐,朱耀麟,韩李婷..基于特征选择与聚类优化的羊绒羊毛分类方法[J].纺织高校基础科学学报,2025,38(5):88-97,10.基金项目
陕西省科技厅面上项目(2024JC-YBMS-491) （2024JC-YBMS-491）
陕西省科技厅自然科学基金重点项目(2023-JC-ZD-33) （2023-JC-ZD-33）
西安市科技局重点产业链技术攻关项目(23ZDCYJSGG0008-2023) （23ZDCYJSGG0008-2023）
