Cooperative system for UAV and ground robot based on cognitive sharing

Since the viewpoint of bottom camera of UAV is greatly different from ground robot, the UAV and ground robot may have a different opinion about same objects. That makes it difficult to describe objects to realize cognitive sharing among multiple robot system contains UAVs and ground robots(heterogeneous robot system). In this paper, we proposed the Multiple Triangle Representation Matching(MTRM) method to realize cognitive sharing between UAV and ground robot through sharing multiple Geometric Relation-based Triangle Representations(GRTR). Our method contributes a robust and effective method for UAV and ground robot identify the same object among similar objects in cooperative objects detecting and carrying task. And we proposed using entropy to evaluate the GRTRs so as to select unique representation. Finally, we demonstrated the effectiveness of proposed method with experiment.