Vision-based tile recognition algorithms for robot grasping task in EAST

Abstract During the operation of the experimental advanced superconducting tokamak (EAST), tiles installed on the vacuum vessel break and fall onto the divertor. To ensure high experimental performance, a robot removes the broken tiles and replaces them with fine tiles. However, one of the key issues for this task is to distinguish fallen tiles from other tiles for grasping. In addition, discrete spots caused by burn and corrosion on the tiles’ surface also bring challenges in vision-based recognition. Here, a novel breakpoint detection algorithm is proposed that recognizes fallen tiles. Experiments are performed to demonstrate the recognition and catch processes of fallen tiles and to test the robustness of the algorithm.