As part of the master’s thesis we developed a program with which we can calculate the corrected grinding path from the point cloud with the help of a given grinding trajectory, where the robot inaccuracies are taken into account. In the initial chapters we explained theoretical backgrounds of the robot, triangulation sensor and 3D scanning with which we help us calculate grinding trajectory. Following this we describe robot system and laser profilometer that we used in the grinding and scanning process. Here we also present how we detected grooves in the point cloud and how we corrected the trajectory. In the results, we showed how point cloud of the product looks before and after processing and the associated 3D model. We found out that some parameters, such as the density of the points, the size of the region of interest, and the inclination of the normal to the Z axis significantly, affect the successful detection of the groove. The calculated trajectory corrections deviate on average by about 1 mm in the Y and Z directions from the programmed one, and in extreme cases by 2 mm in the individual directions.
|
