基于机器视觉的重轨表面缺陷检测系统的研究  

The research of the defective detection system of surface of heavy rail on the basis of the machine visual

作　　者：叶苏茹[1] 胡治民[2] 欧阳奇[1] 李书新[1] 

机构地区：[1]重庆大学机械工程学院,重庆400044 [2]湖南省娄底职业技术学院机电工程系,娄底417000

出　　处：《现代制造工程》2007年第8期89-93,共5页Modern Manufacturing Engineering

摘　　要：针对目前几种常用的重轨缺陷检测方法检测速度慢、精度低等问题,提出一种基于机器视觉的重轨表面缺陷检测系统的设计方法。针对重轨表面常见缺陷种类,根据实验情况对该检测系统的硬件系统进行详细设计。并分别对重轨表面字符和缺陷采用信息定位和模式识别实现图像识别,使重轨缺陷和重轨号等字符信息的相应关系达到自动判别,从而实现缺陷检测的目的。Several kinds of commonly used defective detection methods of heavy rail at present have the problems such as the slow-paced, low grade precision and so on, have proposed a kind of design of the defective system of surface of heavy rail based on the machine vision. To the common defective kinds of surface of heavy rail, have designed the hardware system of this detection system in detail according to the experiment situation. And adopt the information orientation and pattern-recognition to rail surface character and defect separately that realize the pattern recognition, making the corresponding relation in character information such as rail defect and rail number differentiate automatically, thus realize the purpose of the defect detection.

关 键 词：机器视觉 重轨检测 硬件设计 字符识别和缺陷检测 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]