This work presents an alternative approach to the Next-Best-View (NBV) problem for UAVs solely equipped with an RGB camera and with limited a priori knowledge of their target. The vehicle first performs a predefined search routine to locate and obtain several initial views of the target. These images are processed to estimate camera poses, generate a sparse point cloud, and define the 3D bounding volume using the You Only Look Once (YOLO) framework. Subsequent viewpoint selection is informed by optimizing an information gain heuristic at a set of candidate viewpoints surrounding the bounding volume. The point cloud and bounding volume are updated incrementally as the vehicle obtains more information about its surroundings, and the optimal trajectory is recalculated. This approach is validated in a test case.