Visual localization on a multirotor drone - cameras only
Data capture from a small quadcopter drone with two GoPro cameras attached. No IMU, no lidar, no GPS/GNSS, and no other sensors were used. The video shows how we use image features for both visual odometry, to build a map of the environment, and to localize within the map. You can see the blue crosses in the video identifying various features; cross size scales with image patch size. The map view on the bottom shows keyframes being created and connected in a covisibility graph, along with the point cloud of the local sparse map currently being tracked in the imagery. The keyframes in the local map have their little camera pairs drawn above them.