摘要
设计了动态阈值分割算法和缺陷区域提取算法,对钢轨表面掉块、表面裂纹两类典型缺陷图像进行处理,可以准确提取缺陷区域,标定缺陷位置,并统计缺陷特征。搭建了在线钢轨探伤模拟平台,通过高速线阵相机和辅助光源获取图像,由千兆网传输到工控机中,利用Halcon和Visual Csharp编写上层图像处理软件进行在线检测。模拟钢轨探伤实验可以在最快100 km/h速度下,准确的发现钢轨样品表面宽度1 mm的裂纹缺陷,并记录缺陷所在位置。
Designed dynamic threshold algorithm and flaw region extraction algorithm for processing two typical rail surface defects,rail head spalling and cracks.Flaw region can be extracted,while its position and feature were saved and calculated.Built the simulated rail detecting platform,captured images with help of high-speed line-scan camera and illuminants,then transported them to host computer via Gigabit Ethernet.The image-processing software was coding by halcon and visual Csharp.In simulated experiment,the cracks whose width is 1mm can be founded ac-curately and their position will be recorded too,at the speed of 100km/h.
出处
《电子测量与仪器学报》
CSCD
2010年第11期1012-1017,共6页
Journal of Electronic Measurement and Instrumentation
基金
国家"863"计划(编号:2007AA11Z118)资助项目
国家自然基金(编号:60776831)资助项目
关键词
钢轨探伤
表面缺陷
图像处理
特征识别
探伤作业模拟
Rail detecting
surface defect
image processing
feature recognition
simulated rail detecting