Unmanned aerial vehicle (UAV) technology are widely utilized, but in a non-satellite navigation environment or a weak satellite navigation environment, the reliability of unmanned aerial vehicles drops severely, and this is especially the case for swarms. In order to achieve swarm navigation under the condition of no satellite navigation, this paper have developed a small but fully autonomous UAV based on visual sensors. Then, aiming at the problem that the positioning error continuously increases when only relying on visual positioning during the flight of multiple UAVs, this paper proposes an inter-aircraft positioning mutual correction algorithm based on Ultra Wideband (UWB) ranging. In a constrained environment, the stable maintenance of the swarm formation can also be achieved. A variety of real-world field experiments have proved the scalability of our system and the feasibility of the inter-aircraft correction technology.