基于机器视觉的纺纱锭子三维振动检测技术  

3-D vibration detection technology of spinning spindle based on machine vision

作　　者：李阳 陈家新[1] 郑旭栋 唐思凡 孙英健 LI Yang;CHEN Jiaxin;ZHEN Xudong;TANG Sifan;SUN Yingjian(College of Mechanical Engineering,Donghua University,Shanghai 201620,China)

机构地区：[1]东华大学机械工程学院,上海201620

出　　处：《东华大学学报（自然科学版）》2022年第5期101-107,共7页Journal of Donghua University(Natural Science)

摘　　要：针对现有纺纱锭子振动检测系统成本高、适应能力差,以及锭子因轴向窜动难以测量的问题,提出基于单目视觉的锭子三维振动检测技术。设计一套图像采样方案提高系统采样帧率,并基于Canny边缘与RANSAC算法优化检测精度,提出基于空间圆特征与锭子轴线约束的位置求解算法,实现锭子端面三维位置求解。研制样机进行实际的检测,结果表明:检测系统可实现锭子的纵向振动检测,并且分辨率可达2.5μm,采样帧率最高可达4 761帧/s,满足锭子振动检测需求。Aiming at the problems of high cost, poor adaptability and inability to detect longitudinal movement of spinning spindle, a spindle vibration detection method based on monocular camera was proposed. A set of image sampling scheme was designed to improve the sampling frame rate of the system. The detection accuracy was optimized based on the Canny edge and RANSAC algorithm. A pose solving algorithm based on the spatial circle features and the spindle axis constraints was proposed to solve the three-dimensional pose of the spindle end face. After the development of the prototype and the actual testing, the system can detect the longitudinal vibration of the spindle, with the resolution of 2.5 μm and the sampling frame rate of 4 761 frame/s, which meets the requirements of vibration measurement of the spindle.

关 键 词：纺纱锭子 振动检测 机器视觉 空间圆特征 

分 类 号：TS103.811[轻工技术与工程—纺织工程]