In this paper, a Secure A*(SA*) path planning method based on a dual resolution 2.5D layered raster map is proposed to solve the problem of safe path planning for mobile robots in uneven terrain. Firstly, a dual-resolution 2.5D hierarchical grid map is designed, which uses the dual-resolution grid to store obstacle information and elevation information in the environment for saving the storage space of the map. Then, combined with the movement capabilities of the mobile robot, the elevation information in the environment is converted into a constraint factor. And considering the size of the mobile robot, the distance between the mobile robot and the target point is taken as an adaptive factor, and the cost function of the A* algorithm is introduced to ensure that the path of the mobile robot in uneven terrain conforms to its motion ability. Finally, the results of the simulation experiment show that compared with the traditional A* algorithm under the 3D raster map, this scheme can reduce the map storage space by 53.8% and the path elevation standard deviation by 63.9%, which can effectively ensure that the robot is able to be safe in uneven ground.