基于形态学算法的2219铝合金钨极氦弧焊熔池图像特征提取
|
摘要
根据2219铝合金氦弧焊特点,构建了焊接过程实时视觉传感系统,获取了熔池清晰图像;为获得2219铝合金氦弧焊熔池尺寸参数,实现对熔池特征的分析,采用了形态学算法,提取了典型熔池图像边缘形状,并获得了熔池宽度特征尺寸;通过试验验证,提取的熔池图像宽度特征能够准确反映出熔池尺寸信息及变化趋势。研究结果表明:形态学算法准确度高、计算速度快,满足2219铝合金氦弧焊熔池图像实时处理需求,不同阈值及结构元素对图像特征提取效果均会产生较大影响,研究阈值自适应算法,可以满足工程上焊缝熔池特征实时视觉监测的复杂需求。
According to helium-arc welding characteristics of 2219 aluminum alloy,a real-time visual sensing system for welding process was constructed,a clear image of the molten pool was obtained. In order to obtain the size parameters of 2219 aluminum alloy helium arc welding pool and to analyze the characteristics of the molten pool,morphological algorithm was adopted. The edge shape of typical molten pool image was extracted and the characteristic dimension of width of molten pool was obtained. The results show that the width feature of the extracted image can reflect the weld size information accurately. The morphological algorithm has high accuracy and fast calculation speed,which can meet the real-time image processing requirements of 2219 aluminum alloy helium arc welding pool.However,different thresholds and structural elements have great influence on image feature extraction.The threshold adaptive algorithm can meet the complex requirements of real-time visual monitoring of weld pool characteristics in engineering.
According to helium-arc welding characteristics of 2219 aluminum alloy,a real-time visual sensing system for welding process was constructed,a clear image of the molten pool was obtained. In order to obtain the size parameters of 2219 aluminum alloy helium arc welding pool and to analyze the characteristics of the molten pool,morphological algorithm was adopted. The edge shape of typical molten pool image was extracted and the characteristic dimension of width of molten pool was obtained. The results show that the width feature of the extracted image can reflect the weld size information accurately. The morphological algorithm has high accuracy and fast calculation speed,which can meet the real-time image processing requirements of 2219 aluminum alloy helium arc welding pool.However,different thresholds and structural elements have great influence on image feature extraction.The threshold adaptive algorithm can meet the complex requirements of real-time visual monitoring of weld pool characteristics in engineering.
引文
[1]王金凤.基于数学形态学的彩色图像边缘检测[D].秦皇岛:燕山大学,2010.
[2]薛家祥,张丽玲,刘晓. TIG焊视觉图像的二值形态学分析[J].焊接学报,2005,26(6):25-28.
[3]石玗,吴巍,樊丁,等.铝合金MIG焊正面熔池图像视觉传感与处理[J].中国机械工程,2005,16(2):176-179.
[4]王建军,林涛,陈善本.铝合金TIG焊熔池图像的获取与处理[J].机械工程学报,2003,39(5):125-129.
[5]杨家林,高进强,秦国梁,等.基于视觉的激光深熔焊熔池检测及图像处理[J].焊接学报,2011,32(11):21-24.
[6]张强,王正林.精通matlab图像处理[M].北京:电子工业出版社,2009.
[7]梁晓,洪榛,郑凯华,等.基于Open CV的静态场景下实时运动目标检测系统[C].中国自动化学会第二十七届青年学术年会论文集.杭州:中国自动化学会,2012:101-104.
[8]张宇.基于CCD视觉传感的焊缝跟踪计算的研究[D].上海:上海交通大学,2007.
[2]薛家祥,张丽玲,刘晓. TIG焊视觉图像的二值形态学分析[J].焊接学报,2005,26(6):25-28.
[3]石玗,吴巍,樊丁,等.铝合金MIG焊正面熔池图像视觉传感与处理[J].中国机械工程,2005,16(2):176-179.
[4]王建军,林涛,陈善本.铝合金TIG焊熔池图像的获取与处理[J].机械工程学报,2003,39(5):125-129.
[5]杨家林,高进强,秦国梁,等.基于视觉的激光深熔焊熔池检测及图像处理[J].焊接学报,2011,32(11):21-24.
[6]张强,王正林.精通matlab图像处理[M].北京:电子工业出版社,2009.
[7]梁晓,洪榛,郑凯华,等.基于Open CV的静态场景下实时运动目标检测系统[C].中国自动化学会第二十七届青年学术年会论文集.杭州:中国自动化学会,2012:101-104.
[8]张宇.基于CCD视觉传感的焊缝跟踪计算的研究[D].上海:上海交通大学,2007.