Visual guidance of a sealant dispensing robot for online detection of complex 3D-curve seams

Dispensing and seam sealing in the manufacturing industry cause some problems related to human health, time consumption and material waste because of inconsistent sealers. However, at present, in small- and medium-sized factories, manual performance by workers and teaching manipulators by using teach-pendant to recognize seam sealing trajectories, especially for 3D complex seam curves, are still the two main approaches. To improve spraying quality, productivity, and repeatable consistency while reducing cost, this research presents an effective method of using robot and computer vision for autonomous detection of the seam sealing curve. First, the camera calibration methodology is proposed based on the laser triangular algorithm. Then, the coordinates of the seam curve are computed by applying the image processing algorithm. After that, the coordinates are processed and converted to the coordinates of the robot and transmitted to the controller to perform the task. By using this methodology, the robot can perform the gluing task without knowing the trajectory in advance; moreover, the system is not overly complicated and has high accuracy as well as a fast processing speed. The efficiency of the system is proven through experiments on 2D and 3D complex curves.