Camera-Based Homing System

Authors: Ofir Mazar, Gleb Merkulov
Supervisor: Prof. Tal Shima

Goals

  • Design and implement an onboard homing system for a ground robot based on 360-degee FOV camera and inertial robot measurements to extend the laboratory framework.
  • Demonstrate the system performance in the pursuite experiments with ground robots using classical guidance laws.

Description

The use of a global positioning system (e.g. OptiTrack) in the Cooperative Autonomous SYstems (CASY) laboratory is convenient for analyzing the performance of guidance and control algorithms on actual vehicles with complex dynamics but under near-perfect information assumptions. However, the real-world guidance systems rely on the relative pursuer-evader motion information, which is derived from the seeker and the inertial measurements. These measurements are less accurate but better mimic the realistic system setup. This project aims to design a framework that allows target tracking using an onboard measurement system only. A 360-degree Field of View (FOV) camera will be used as the “seeker” to provide bearing (and range) to the target. Subsequent fusion with the inertial measurements will be used to improve and extend the information about the relative motion. The performance of the homing system is to be demonstrated in a pursuit scenario using classical guidance laws. Such extension of the laboratory framework paves way for the experimental validation of integrated guidance and estimation algorithms and using smaller bodies as targets, thus increasing the number of vehicles in the scenario.