In this paper, we propose a novel two-stage optimization-based path planner for the 3D reconstruction of a target scene using a quadrotor UAV (unmanned aerial vehicle). The proposed method solves two-stage optimization problems to find the shortest, safe, and energy-efficient trajectory that covers the target scene. Firstly, the shortest and safe coverage path is obtained by solving the obstacle-aware Traveling Salesman Problem (O-TSP), which visits whole viewpoints. Then, the resulting path is refined as an energy-efficient trajectory by optimizing the cost function that considers the snap of the UAV. The proposed method is validated with a simulation.