Vision-based indoor localization of nano drones in controlled environment with its applications

TL;DR

The paper tackles GPS-denied indoor localization for nano drones by integrating off-board monocular vision with WhyCon markers and a parallel PID control framework, achieving a localization error of $\pm 3.1$ cm at a low cost (~$50$ USD). The authors validate a complete system comprising a WhyCon-based tracker, off-board processing, ROS communication, and multi-formation control, demonstrating autonomous landing on a moving target, 3D path planning in V-REP with OMPL, and multi-drone coordination. The contributions center on a cost-effective, integrated localization-control stack and its application to education and basic research, with openly available ROS packages. The work highlights practical implications for affordable indoor robotics labs and sets a foundation for scaling with additional monocular cameras and wider controlled spaces.

Abstract

Navigating unmanned aerial vehicles in environments where GPS signals are unavailable poses a compelling and intricate challenge. This challenge is further heightened when dealing with Nano Aerial Vehicles (NAVs) due to their compact size, payload restrictions, and computational capabilities. This paper proposes an approach for localization using off-board computing, an off-board monocular camera, and modified open-source algorithms. The proposed method uses three parallel proportional-integral-derivative controllers on the off-board computer to provide velocity corrections via wireless communication, stabilizing the NAV in a custom-controlled environment. Featuring a 3.1cm localization error and a modest setup cost of 50 USD, this approach proves optimal for environments where cost considerations are paramount. It is especially well-suited for applications like teaching drone control in academic institutions, where the specified error margin is deemed acceptable. Various applications are designed to validate the proposed technique, such as landing the NAV on a moving ground vehicle, path planning in a 3D space, and localizing multi-NAVs. The created package is openly available at https://github.com/simmubhangu/eyantra_drone to foster research in this field.

Figures (15)

• Figure 1: Comparison between flight time, drone weight, and cost for localization of nano aerial vehicle (NAV), miniature aerial vehicle (MAV) and tactical unmanned aerial vehicle (TUAV). Using NAV along with WhyCon marker for localization, substantially reduces the setup cost.
• Figure 2: Coordinate frames of the NAV and overhead camera. WhyCon marker, shown with an inner white circle and black boundary, is mounted on top of the NAV. Changes in roll ($\phi$), pitch ($\theta$), and throttle result in movement along the y, x, and z axes respectively.
• Figure 3: Block diagram of external PID-Controller used for roll ($\phi$), pitch ($\theta$), and throttle. These controllers work in parallel to stabilize the drone.
• Figure 4: Arrangement of a custom-controlled localization environment for NAV utilizing an adapted WhyCon system.
• Figure 5: $\phi$, $\theta$ and Altitude error with internal control. As depicted in the graph, there is an evident increase in error of roll, pitch, and altitude with time. This indicates the necessity for an external controller.