Semantical obstacle representation model for fast UAV path planning

Unmanned aerial vehicles (UAVs) path planning has received significant attention from both military and academia domains, current researches are more focus on how to implement efficient path planning algorithms, but ignored the prime effects of planning environment and obstacles. In this paper, authors proposed a novel multi-granularity semantic-based obstacles model, and abstracting the constraints and obstacles in the planning environment as different constraint agents. On this basis, authors adopted the backtracking searching framework between constraint agents to achieve the feasible results, and the corresponding simulation towards multi-UAVs path planning and the path checking algorithm are implemented.