Prototype of a Computer Vision-Based CubeSat Detection System for Laser Communications

Up to now, CubeSat nano-satellites have strong limitations in communication data rates (∼100\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sim \hbox {100}$$\end{document} kbps) and bandwidth due to the strictness of CubeSat standard. However, if they could be endowed with optical communications (data rates up to 1 Gbps in optimal state), CubeSat applications would exponentially increase. Nonetheless, laser communications face some important drawbacks as the development of a very strict and accurate tracking mechanism. This work proposes an on-board fine pointing system to locate an optical ground station beacon using an embedded system complying with the restrictive CubeSat standard. Such on-board fine pointing system works based on computer vision. The experimental prototype is implemented in Matlab/Simulink, within a Raspberry Pi 3B. The main outcome is the usage of off-the-shelf components (COTS), obtaining an efficient tracking with low power consumption in very noisy and reflective environments. The developed system proves to be fast, stable and strong. It also satisfies the strict size and power consumption restrictions of CubeSat standard.