Fast and Omnidirectional Relative Position Estimation With Circular Markers for UAV Swarm

Swarms of unmanned aerial vehicles (UAVs) have been applied in various areas. An essential prerequisite for a swarm is the relative position among individuals. Although the Global Navigation Satellite System (GNSS) provides an effective approach to acquiring the position of each member, it fails in many scenes, such as indoors. Vision-based methods are essential substitutes for GNSS in GNSS-denied scenarios. Considering the lack of a line marker for vision-based methods, an infrared circular marker is designed for the first time, and a geometry model is developed to estimate the relative position. Then, a clustering model is adopted to separate different markers in the image, considering the application in swarms. The relationship between the relative distance and the pixel intensity is introduced to provide a reference when occlusion happens. Finally, four experiments are performed to show the accuracy and efficiency of this method in GNSS-denied environments. In addition, the method is robust to partial occlusion by estimating relative distance through pixel intensity and the combination of the clustering model and geometry model.