基于CCD图像处理的焊缝识别技术研究  

Research on weld seam recognition technology based on CCD image processing

作　　者：丛焕武 郭福娟[2] 吕飞[2] 王倩[2] 

机构地区：[1]哈尔滨华崴焊切股份有限公司,哈尔滨150086 [2]哈尔滨工业大学电气工程及自动化学院,哈尔滨150001

出　　处：《电子测量技术》2012年第3期73-77,共5页Electronic Measurement Technology

基　　金：国家自然科学基金(60901042);国家自然科学基金(61171189)资助项目

摘　　要：在分析焊缝的CCD数字图像信息的基础上,提出了优化的图像处理和焊缝提取算法,有效地实现了焊缝的识别。对图像进行预处理,设计了扩展的自适应中值滤波器以去除噪声和干扰;提出了改进的自适应阈值分割法(Otsu)对图像进行二值化,得到二值图像;选择Roberts算子识别焊缝边缘信息,并采用Hough变换获取焊缝中心线,得到焊缝的坐标值。在VC++平台上分析和处理焊缝图像,实现上述算法,结果表明,所设计算法不仅能够快速、准确地识别焊缝,而且具有计算量小、实时性强的优点,有很高的理论和应用价值。On the basic of the CCD digital image information of weld seam,an optimized image processing and seam feature extraction algorithm is proposed,which has efficiently achieved weld seam recognition.The expanded adaptive median filter is designed to remove the disturbance and noise in the image.The adaptive threshold selection image segmentation method is modified to compress the image data and realize image binaryzation.The Roberts operator is selected to detect the welding edge and the axis of the seam is obtained through Hough transform.The experiment is implemented on the VC＋＋ platform to process the image data and execute the algorithm.The results show that the proposed algorithm can recognize the weld seam accurately and quickly and it also has the characteristics of less calculation and high real-time,which means high theoretical and application value.

关 键 词：焊缝识别 自适应中值滤波 图像分割 边缘检测 HOUGH变换 

分 类 号：TP2[自动化与计算机技术—检测技术与自动化装置]