基于人工智能的广西都安山羊皮革瑕疵自动检测分级研究

Research on Automatic Detection and Grading of Imperfections in Du'an Goat Leather from Guangxi Province,based on Artificial Intelligence

通过文献研究和案例分析,文章探究了基于人工智能的皮革瑕疵自动检测分级技术。文章重点探究如何解决广西都安山羊板皮经脱毛鞣制后,由于其本身的皮质特性使得针对其进行人工检测的人员因受到天气条件、光线变化、视力健康等因素干扰而导致其在标记瑕疵时经常出现错检或漏检等出错率较高的问题。研究认为,基于深度学习(DL)中的全卷积网络(FCN)和机器学习(ML)中的深度神经网络(DNN)等人工智能技术能够对广西都安山羊皮革各类瑕疵进行有效地自动检测分级,最大程度上使广西都安山羊皮革能够更好地对标符合当前中国-东盟RCEP(区域全面经济伙伴关系协定)中对于成员国皮革进出口贸易规范(质量标准)的要求。

Through literature review and case analysis,the paper explores the technology of AI-based automatic defect detection and grading in leather.It focuses on addressing the challenges faced by manual inspectors of Du'an goat leather,where factors such as weather conditions,variations in lighting,and visual impairments often lead to high error rates in defect marking.The study argues that artificial intelligence techniques such as Fully Convolutional Networks(FCN)in deep learning(DL)and Deep Neural Networks(DNN)in machine learning(ML)can effectively automate the detection and grading of various defects in Du'an goat leather.This approach maximizes compliance with the quality standards required under the China-ASEAN RCEP(Regional Comprehensive Economic Partnership)for leather imports and exports among member countries.