基于图像处理技术的工件长度在线测量  

Online Measurement of Workpiece Length Based on Image Processing Techniques

作　　者：沙毅[1] 曹英禹[1] 王经武 盛国松 

机构地区：[1]东北大学信息科学与工程学院,辽宁沈阳110004 [2]鞍山挪洋科技有限公司,辽宁鞍山114044 [3]辽宁省鞍山地震台,辽宁鞍山114001

出　　处：《东北大学学报（自然科学版）》2005年第10期957-959,共3页Journal of Northeastern University(Natural Science)

基　　金：国家高技术研究发展计划项目(2002AA784030)

摘　　要：针对轧钢厂加热钢锭传输线,提出了一种工件长度在线测量方法.利用CCD摄像机对生产线上工件的图像信息进行远距离采集.考虑到实时在线检测要求算法简单计算速度快,在对图像信息进行边缘噪声滤除的基础上,利用图像灰度的梯度算子提取工件边缘特征,并进行边缘识别,从而实现工件长度的在线测量.为补偿摄像镜头的非线性影响,进行了二次多项式拟合处理,以提高测量精度.同时,介绍了实时在线测长系统的组成和功能,并对其测量速度和精度进行了实际测试分析,结果表明,被测工件长度在400～1200mm之间,运行速度不高于5m/s时,测量误差低于0.4%.Aiming at a conveyer line of heated ingots in a roll mill, a method is presented for online measurement of workpiece length. The image information of a workpiece on a production line is acquired by a CCD camera at distance, and then the edge noise of the image is filtered. The real-time measurement is carried out using a simple algorithm or the gradient operator of image grey ,scale to detect and recognize quickly the edge of an image, thus acquiring the workpiece length. The quadratic polynomial is fitted to compensate the non-lincarily of the camera lens ,so as to improve the measuring precision. Introduces the structure and functions of a real-time system for online measurement and analyzes its speed and precision. The results show that when the workpiece length is 400-1200 mm with conveying speed less than 5 m/s, the measuring error is lower than 0.4 %.

关 键 词：在线测量 自动检测 图像处理 边缘检测 摄像机 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]