基于双目视觉的焊缝跟踪系统研究
|
摘要
在自动焊接技术中,需要焊枪对焊缝实时而准确地跟踪。单目视觉系统不能实现对曲面焊缝的焊接。引入双目视觉技术,利用视觉反馈,实时监控焊点位置,减小自动焊接工作过程中的焊接误差,提高焊缝精度;对双目视觉系统进行标定,得到相机内外参数,利用投影矩阵M得到焊点三维坐标,并以此计算焊点的理论坐标误差,通过实时补偿,满足高精度焊接要求。实验表明,在加装双目视觉传感系统后,焊缝平均精度可达0.68mm,满足自动焊接的要求。
In automatic welding technology,weld joint needs to be tracked by the welding gun accurately in real time.The monocular vision system can not realize the welding of curved weld. In this paper,binocular vision technology is introduced and visual feedback is used to monitor the position of welding spot in real time,which reduce welding error in automatic welding process and improve weld accuracy;the binocular vision system is calibrated to obtain the internal and external parameters of the camera,and the 3 d coordinates of the solder points are obtained by using the projection matrix M. Then the theoretical coordinate error of weld joints is calculated,and the high precision welding requirements are met through real-time compensation. The experimental results show that the average welding accuracy can reach0.68 mm and meet the requirements of automatic welding after the binocular vision sensor system is installed.
In automatic welding technology,weld joint needs to be tracked by the welding gun accurately in real time.The monocular vision system can not realize the welding of curved weld. In this paper,binocular vision technology is introduced and visual feedback is used to monitor the position of welding spot in real time,which reduce welding error in automatic welding process and improve weld accuracy;the binocular vision system is calibrated to obtain the internal and external parameters of the camera,and the 3 d coordinates of the solder points are obtained by using the projection matrix M. Then the theoretical coordinate error of weld joints is calculated,and the high precision welding requirements are met through real-time compensation. The experimental results show that the average welding accuracy can reach0.68 mm and meet the requirements of automatic welding after the binocular vision sensor system is installed.
引文
[1]陈善本,吕娜.焊接智能化与智能化焊接机器人技术研究进展[J].电焊机,2013,43(05):28-36.
[2]柳刚,李俊岳,李恒,等.焊接电弧光谱的分布特征[J].机械工程学报,2000,36(5):58-61.
[3]徐杰.机器视觉中摄像机标定Tsai两步法的分析与改进[J].计算机工程与科学,2010,32(04):46-47.
[4]姜雨彤,杨进华,刘钊,等.双目CCD测距系统的高精度标定[J].计算机工程,2013,39(07):228-232.
[5] Juan L,Gwan O.A Comparison of SIFT,PCA-SIFT and SURF[J].International Journal of Image Processing,2010,3(04):143-152.
[6] Vallerand S,Kanbara M,Yokoya N. BinocularVision-based Argume-nted Reality System with an Increased Registration Depth Using Dynamic Correction of Feature Positions[C].Virtual Reality,IEEE,2003.
[7]邹凤娇,苏显渝,李美菊.基于共面点的摄像机线性标定法[J].光电工程,2005,32(04):72-73.
[8]肖心远,石永华,王国荣,等.机器人水下焊缝跟踪中双目立体视觉系统[J].焊接设备与材料,2009,38(01):37-40.
[9]郑鹭斌,王晓栋,严菲.一种基于线结构光的焊缝三维重建方法[J].激光与光电子学进展,2014,51(04):122-128.
[2]柳刚,李俊岳,李恒,等.焊接电弧光谱的分布特征[J].机械工程学报,2000,36(5):58-61.
[3]徐杰.机器视觉中摄像机标定Tsai两步法的分析与改进[J].计算机工程与科学,2010,32(04):46-47.
[4]姜雨彤,杨进华,刘钊,等.双目CCD测距系统的高精度标定[J].计算机工程,2013,39(07):228-232.
[5] Juan L,Gwan O.A Comparison of SIFT,PCA-SIFT and SURF[J].International Journal of Image Processing,2010,3(04):143-152.
[6] Vallerand S,Kanbara M,Yokoya N. BinocularVision-based Argume-nted Reality System with an Increased Registration Depth Using Dynamic Correction of Feature Positions[C].Virtual Reality,IEEE,2003.
[7]邹凤娇,苏显渝,李美菊.基于共面点的摄像机线性标定法[J].光电工程,2005,32(04):72-73.
[8]肖心远,石永华,王国荣,等.机器人水下焊缝跟踪中双目立体视觉系统[J].焊接设备与材料,2009,38(01):37-40.
[9]郑鹭斌,王晓栋,严菲.一种基于线结构光的焊缝三维重建方法[J].激光与光电子学进展,2014,51(04):122-128.