Research on UAV Delivery Planning Based on Deep learning and Visual SLAM

In recent years, with the rising demand for takeaways, customers' requirements for takeaway delivery efficiency have gradually increased. For this problem, a method of using UAVs as carriers for takeaway delivery is proposed, because UAVs do not need to consider traffic problems, the takeaway delivery time can be greatly shortened, and UAVs can be recycled and reused, and the distribution cost is also reduced. In this paper, we will use visual SLAM to simulate the delivery environment, and plan the path and calculate the optimal path based on deep learning reference A* algorithm and ant colony algorithm. Finally, the feasibility of the method is verified.