A method and apparatus for performing two-dimensional video alignment onto three-dimensional point clouds. The system recovers camera pose from camera video, determines a depth map, converts the depth map to a Euclidean video point cloud, and registers two-dimensional video to the three-dimensional point clouds.