A Review of Autonomous UAV Landing Technologies in Uncooperative Environments

The rapid advances in unmanned aerial vehicle (UAV) technologies have expanded their military, emergency, logistical, and environmental monitoring applications, intensifying demands for autonomous operation in complex dynamic environments. Autonomous landing, as a critical terminal phase of UAV missions, profoundly impacts both mission success and system integrity. In uncooperative scenarios, terrains are unpredictable. There is also no prior information or external support available. In such cases, UAVs must exclusively depend on onboard sensors and real-time decision-making to ensure safe touchdown. Conventional navigation systems—including Global Navigation Satellite Systems (GNSS) and Inertial Navigation Systems (INS)—often prove inadequate due to environmental interference and limited precision in intricate landscapes. Vision-based landing solutions have consequently emerged as promising alternatives, offering advantages in cost efficiency, low power consumption, and operational robustness. This review synthesizes recent progress in visual autonomous landing methodologies, categorizing existing frameworks according to vision system deployment (onboard versus ground-based) and target dynamics (static versus moving platforms). We present a comparative analysis of these approaches, delineate their respective merits and constraints, and critically examine persistent challenges such as dynamic adaptability, real-time computational efficiency, and adverse weather resilience. Future research trajectories should prioritize multisensor fusion architectures, enhanced algorithmic robustness, and operational optimization under extreme conditions to achieve reliable autonomous landings in non-cooperative settings.