A Precision Drone Landing System using Visual and IR Fiducial Markers and a Multi-Payload Camera

TL;DR

A method for autonomous precision drone landing with fiducial markers and a gimbal-mounted, multi-payload camera with wide-angle, zoom, and IR sensors and a high-level control policy to manage initial search for the landing pad and subsequent searches if it is lost is proposed.

Abstract

We propose a method for autonomous precision drone landing with fiducial markers and a gimbal-mounted, multi-payload camera with wide-angle, zoom, and IR sensors. The method has minimal data requirements; it depends primarily on the direction from the drone to the landing pad, enabling it to switch dynamically between the camera's different sensors and zoom factors, and minimizing auxiliary sensor requirements. It eliminates the need for data such as altitude above ground level, straight-line distance to the landing pad, fiducial marker size, and 6 DoF marker pose (of which the orientation is problematic). We leverage the zoom and wide-angle cameras, as well as visual April Tag fiducial markers to conduct successful precision landings from much longer distances than in previous work (168m horizontal distance, 102m altitude). We use two types of April Tags in the IR spectrum - active and passive - for precision landing both at daytime and nighttime, instead of simple IR beacons used in most previous work. The active IR landing pad is heated; the novel, passive one is unpowered, at ambient temperature, and depends on its high reflectivity and an IR differential between the ground and the sky. Finally, we propose a high-level control policy to manage initial search for the landing pad and subsequent searches if it is lost - not addressed in previous work. The method demonstrates successful landings with the landing skids at least touching the landing pad, achieving an average error of 0.19m. It also demonstrates successful recovery and landing when the landing pad is temporarily obscured.

Figures (6)

• Figure 1: The angles defining the control policy for the drone (D) to approach the landing pad (LP). $\phi$ represents the pan, and $\theta$ the tilt, from the drone to the landing pad. $\vec{H}$ represents the heading of the drone, in the direction of the $y$-axis. The landing pad is at the base of the vector $\vec{LP}$, and the vector extends in the direction of the landing pad's yaw. The angle $\psi$ represents the relative yaw from the drone's heading to the landing pad's heading. The altitude $a$ and range $R$ are not used or needed for this method.
• Figure 2: The control policy. For Aim Camera, Aim Drone, and Approach, transition conditions are based on $\Theta_c = \Theta_d = \Theta_a = 3\degree$. For Yaw Alignment, $\Theta_y = 1\degree$. For Horizontal Alignment, $\Theta_h = 2\degree$. For Descent, $Z_\text{min}=2$, and $S_\text{max}=32\%$. For Commit, the flight controller indicates motor stop.
• Figure 3: RGB and pictures of the landing pads, taken from the drone. Top: visual landing pad. Bottom left: passive landing pad (at ambient temperature, high reflectivity). Bottom right: active (heated) landing pad.
• Figure 4: After experimental landing: the Matrice 350, H20T and Raspberry Pi (inside the top-mounted case) on top of the passive landing pad. The landing pad is placed on a backdrop of grass with small patches of snow.
• Figure 5: Reflection of the drone's heat signature in the passive landing pad. Left: 13m altitude. Middle: 10m altitude. Right: 7m altitude.