In order to solve the reconnaissance route planning problem of unmanned aerial vehicle, the memetic algorithm was adopted, which was widely investigated recently. The particle swarm optimization was selected as the main planning method while the simulated annealing algorithm was applied in the local search. For the objective function, many factors were analyzed including the tactical purpose, flying range, security and time. Under the constraint conditions, not only the environment, but also the UAV itself was taken into consideration. The simulation results indicate that the memetic algorithm based on particle swarm optimization and simulated annealing is more effective than the single PSO, especially for the convergence speed.