Abstract: A path planning method based on improved sparrow search algorithm is proposed to solve the problem of reconnaissance mission path planning for unmanned aerial vehicle system. A task model of the air-ground cooperative system with two UAVs mounted by unmanned vehicle was proposed, which took minimizing the time of reconnaissance task as the optimization objective. The solving quality and convergence speed of the algorithm were improved by optimizing and adjusting the position updating strategies of individual population in the sparrow search algorithm. Simulation verifies the effectiveness of the cooperative reconnaissance model and improved sparrow search algorithm. The results show that the task time of the air-ground cooperative reconnaissance model is reduced by 39.5%, and the convergence rate of the improved sparrow search algorithm is increased by 13%.