分析测试学报2026，Vol.45Issue(2)：358-365,8.DOI:10.12452/j.fxcsxb.25070403
基于人工智能的动物毛混纺纤维含量定量分析研究
Study on Quantitative Analysis of Animal Hair Blended Fiber Content Based on Artificial Intelligence
摘要
Abstract
Different from the GB/T 16988-2013 standard"Quantitative determination for mixtures of special animal fibre and wool"which uses manual methods to distinguish and measure the fiber content of animal hair and its blended products,the fiber content analysis technology based on artifi-cial intelligence uses a microscope to automatically scan and capture animal fiber images,integrates advanced methods such as image recognition and image processing,establishes a fiber classification model,and then completes the automatic quantitative analysis of fiber content.This paper introduc-es the principle of the artificial intelligence quantitative analysis method for the composition of animal hair fibers,designs a balance and uniformity level test for animal hair blended fibers with 8 laborato-ries and 4 types and 3 mass levels each,conducts a method validation test for the artificial intelli-gence quantitative analysis method,and evaluates the trueness and precision of the method.The test results show that the accuracy of the artificial intelligence quantitative analysis method is reliable and the precision meets the requirements of detection applications.It can be equivalent to replacing the manual test method of the GB/T 16988-2013 standard,and provide more efficient and stable quanti-tative detection results,and has broad application prospects.关键词
毛混纺纤维/人工智能/定量分析/方法确认Key words
hair blended fiber/artificial intelligence/quantitative analysis/method validation分类
化学化工引用本文复制引用
颜刚华,王文,黄宇斌,刘卓钦,谢飞,杨二涛,黄海敏,兰丽丽,袁志磊,费静..基于人工智能的动物毛混纺纤维含量定量分析研究[J].分析测试学报,2026,45(2):358-365,8.基金项目
海关总署科研项目(2024HK064) （2024HK064）
