An Evolutionary Method for Joint Optimization of UAV Path and Camera Direction Planning with Optional Viewpoints

This paper addresses the Unmanned Aerial Vehicle (UAV) path planning problem with optional viewpoints, formu lated as a joint optimization that considers surface coverage qual ity and flight energy consumption. The problem involves strong interdependence among viewpoint selection, path sequencing, and camera orientation, resulting in a large-scale combinatorial challenge that cannot be efficiently solved by off-the-shelf linear programming solvers. To tackle this, we develop a dual-layer chromosome genetic algorithm (GA) that encodes both UAV paths and camera directions. The GA employs adjacency-preserving edge-recombination crossover, feasibility-aware node replacement mutation, and a coverage repair operator, enabling constraint preserving search while satisfying the sample point coverage requirement. Experiments on urban 3D scenes confirm that the GA consistently produces feasible solutions, achieving near optimal quality on the majority of instances and solving large scale problems intractable for Gurobi with a 94% time reduction, thereby demonstrating superior scalability and efficiency for UAV photogrammetric planning.