花式纱线条干检测的视觉传感器标定方法研究  

Research on the Calibration Method of Visual Sensor in Fancy Yarn Evenness Detection

作　　者：杨承翰 张建鹏 徐云 YANG Chenghan;ZHANG Jianpeng;XU Yun(School of Information Science and Engineering,Zhejiang Sci-Tech University,Hangzhou 310018,China)

机构地区：[1]浙江理工大学信息科学与工程学院,浙江杭州310018

出　　处：《软件工程》2024年第9期56-61,共6页Software Engineering

基　　金：国家自然科学基金项目(62203393);浙江省自然科学基金项目(LQ20F030019)。

摘　　要：针对机器视觉的检测系统存在相机畸变、受环境光照影响等问题,提出一种花式纱线条干检测的视觉传感器标定方法。构建了花式纱线条干检测系统样机,对相机内参和畸变系数进行标定。提出图像预处理和图像降维的方法捕获用于表征纱线条干的宽度序列,解决了系统受环境光照影响的问题。实验结果表明,标定后的检测系统能够有效抑制相机畸变的影响,标定后的纱线平均直径测量值比未标定的提高了4.48%,标定后的纱线条干变异系数比未标定的降低了8.92%。该方法可为纱线条干检测的实际应用提供参考。Aiming at the issues in machine vision detection systems such as camera distortion and environmental lighting influence,a visual sensor calibration method for detecting fancy yarn evenness is proposed.A prototype of the fancy yarn evenness detection system is constructed,and the camera internal parameters and distortion coefficients are calibrated.Methods of image preprocessing and dimension reduction are introduced to capture the width sequence characterizing the yarn evenness,addressing the problem of environmental lighting influence on the system.Experimental results show that the calibrated detection system effectively suppresses the impact of camera distortion,with the average diameter measurement of the calibrated yarn increasing by 4.48%compared to the uncalibrated system,and the coefficient of variation of the calibrated yarn evenness decreasing by 8.92%compared to the uncalibrated ones.The proposed method can provide a reference for the practical application of yarn evenness detection.

关 键 词：花式纱线 条干检测 相机畸变 标定 机器视觉 图像处理 宽度序列 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]