摘要
通过视觉传感器获取焊缝图像,采用多种去噪技术对焊缝图像中的噪声进行滤波试验。针对焊缝图像中的干扰噪声特性,对各种滤波方法进行了分析和比较,并对滤波效果较好的焊缝图像进行边缘检测计算。提出3种识别焊缝特征的组合图像处理方法,并研究提取焊缝中心特征信息的效果和精度。试验结果表明,对焊缝图像采用Wiener滤波与Robert边缘检测算法能较好地提取焊缝中心位置的特征信息。
The experiments were performed to study the technologies of eliminating the noises of seam images obtained from a visual sensor.Some filter methods were analyzed to find the characteristics of the weld image noises.The weld edge coordinates were calculated for those effective filtering weld images.Three image integrated approaches for detecting the weld characteristics were presented.Also,the efficiency and precision of detecting the seam center of those methods were compared.The experimental results show that the method which adopts Wiener filter and Robert edge detection is the most efficient to obtain the accurate information of the seam center.
出处
《电焊机》
2005年第2期58-61,共4页
Electric Welding Machine
基金
国家自然科学基金资助项目(60375012)
广东省自然科学基金资助项目(020176)
关键词
焊缝图像
滤波
焊缝中心
特征提取
检测精度
weld image
filtering
weld center
character extraction
detection precision€
