On-line robotic spray painting using machine vision

In the present work, a machine vision system was used to guide a general 6R robot along a planar path for spray painting of flat polygonal objects. The vision system provides an intelligent interface to calculate the spray painting path for new and unknown parts. The image of the polygonal object was acquired by the vision system and digitized for further processing. The object was segmented from the surroundings and an efficient algorithm was used to decipher the coordinates of the vertices of the polygonal object. These coordinates form the input to a path generation software, which in turn uses an algorithmic method to generate an optimal path. The method of `staircasing' was used to generate the path. Once the end effector path was calculated, a novel iterative approach is used to solve the inverse kinematics of a general 6R manipulator. The joint angles calculated serve as the robot controller to direct the robot end effector along the prescribed path.