基于视觉伺服的羊毛衫自动对边方法研究  

Research on automatic edge alignment method for woollen sweater based on visual servo

作　　者：孙国栋[1] 徐威[1] 宋丽萍 

机构地区：[1]湖北工业大学机械工程学院,湖北武汉430068 [2]南京普爱射线影像设备有限公司,江苏南京211112

出　　处：《毛纺科技》2014年第2期51-53,共3页Wool Textile Journal

基　　金：国家自然科学基金项目(项目编号:51205115;51075130);湖北省教育厅科学研究计划资助项目(项目编号:Q20121408)

摘　　要：针对当前毛针织产品需求量不断增大,而我国的毛针织产品整体质量较差,缝纫设备落后的现状,提出了一种基于视觉伺服的羊毛衫自动对边方法。该方法采用机器视觉技术把采集的面料彩色图像转化为黑白图像,设置ROI并通过中值滤波、阈值分割、图像腐蚀后,搜索刻度线区域轮廓的最小包络矩形进行面料边沿提取,计算面料边沿至基准线的偏差,进一步转化为步进电动机的脉冲信号,驱动丝杆调节机构实现上下两片面料边沿的对齐。通过大量实验验证,该自动对边方法生产的羊毛衫缝纫质量好,且生产效率高。With the increasing demand for the current wool knitted products and the overall poor quality of wool knitted products and backward present situation of sewing equipment in our country, an automatic edge alignment method for woollen sweater is proposed based on visual servo. The method adopts machine vision technology to translate color images of the fabric into grey images, set the ROI and through median filtering, threshold segmentation, and image corrosion, search the minimum bounding rectangle of scale line contours, extracts the fabric edge from collected images, calculates the deviation between the fabric edge and the baseline, and transforms further the deviation into the pulse signal of stepper motor, finally drives screw regulating mechanism to achieve edge alignme two pieces of fabric. A large number of experiments have validated the proposed m quality of woollen sweaters and production efficiency.

关 键 词：机器视觉 羊毛衫 缝合 自动对齐 

分 类 号：TS941.63[轻工技术与工程—服装设计与工程]