图像分割方法在焊缝跟踪过程中的应用  

Application of image segmentation method in seam recognition process

作　　者：赫健[1] 王克成[1] 朱连成[1] 李玥[2] 王鸿颖[3] 

机构地区：[1]辽宁科技大学电子与信息工程学院,辽宁鞍山114051 [2]中钢集团鞍山热能研究院有限公司,辽宁鞍山114001 [3]鞍钢股份有限公司能源动力总厂,辽宁鞍山114001

出　　处：《辽宁科技大学学报》2013年第1期33-37,共5页Journal of University of Science and Technology Liaoning

摘　　要：为提高焊缝超声波探伤的实时跟踪,提出将图像分割方法应用于白线跟踪法和激光跟踪法两种不同焊缝跟踪系统中。一种是采用Hough变换边缘检测法对检测图像中的目标边缘进行提取,另一种是利用区域增长法对检测图像中的目标区域和背景区域进行有效区分,这两种方法都能很好地提取出图像中的目标特征信息。仿真结果表明,两种图像分割算法在焊缝跟踪过程中能够准确地识别出焊缝的特征信息,确保跟踪系统的实时性。In order to improve the real-time tracking of seam ultrasonic inspection,two image segmentation methods are applied to two different seam tracking systems,which are white line tracking and laser tracking.In the first method,the target edge in inspecting image is extracted by edge detection of Hough transform.In the second method,the target edge and background in inspecting image are effectively distinguished by region growing.The target characteristic information in image is better extracted by two methods.The simulation results indicate that both methods can accurately recognize the seam characteristic information,and ensure the real-time in tracking system.

关 键 词：焊缝跟踪 图像分割 HOUGH变换 区域增长 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]