With the rapid development of unmanned vehicle technology, its applications in military, civilian, and commercial sectors are becoming increasingly widespread. Alongside the evolution of autonomous systems for unmanned vehicles, ensuring precise autonomous navigation has become an important and complex challenge. This project aims to develop a low-cost autonomous navigation system for unmanned vehicles in unknown environments. The system will utilize a monocular camera and computing unit, combined with advanced monocular depth estimation technology, the A* navigation algorithm, and Model Predictive Control (MPC) algorithm. By capturing environmental images with the monocular camera and generating real-time 3D maps of the surrounding environment using depth estimation algorithms, the system will plan safe and efficient driving paths through the integration of the A* algorithm. Finally, the MPC model predictive algorithm will be employed for trajectory tracking control, enabling autonomous navigation and exploration in complex environments such as GNSS-denied areas. Simulation experiments indicate that our research demonstrates exceptional capabilities in real-time environmental perception and dynamic path planning.