基于机器学习的布匹瑕疵识别研究

纺织工业是我国制造业出口的重要组成部分。布匹的质量控制在纺织工业中尤为重要,而布匹瑕疵是影响布匹质量控制的重要因素之一。在中小企业中,布匹瑕疵识别主要依靠人工流水线作业,存在着人工成本高、人眼识别准确度低等问题。因此,一个有效的布匹瑕疵检验系统是十分必要的,布匹瑕疵分类算法是保证疵点判决效率的核心。基于布匹生产企业存在的问题,有针对性地研究了机器学习与计算机视觉的布匹瑕疵识别算法的基本原理,介绍了各类布匹瑕疵识别中的检测与分类算法,将最近发展迅速的机器学习的理论研究引入布匹瑕疵识别中,对涉及机器学习的模式识别算法进行了介绍。

Study on fabric defect detection based on machine learning

Textile industry is an important part of China’s manufacturing export.The quality control of fabric is very important in textile industry.The defect of fabric is one of the important factors affecting the quality control of fabric.At present,in small and medium-sized enterprises,fabric defect identification mainly depends on manual assembly line,which has problems such as high labor cost and low accuracy of human eye recognition.Therefore,an effective fabric defect inspection system is very necessary.Fabric defect classification algorithm is the core to ensure the efficiency of defect judgment.Based on the problems existing in fabric manufacturing enterprises,this paper studies the basic principle of fabric defect recognition algorithm based on machine learning and computer vision,introduces the detection and classification algorithm of various kinds of fabric defect recognition,and introduces the theoretical research of machine learning into fabric defect recognition.This paper introduces the model recognition algorithm involved in machine learning.