HSV color space based buoy detection module for autonomous underwater vehicles

Abstract

In this paper, we describe the development and the testing process of an underwater buoy detection module (package) written for the Robot Operating System (ROS). We developed this module for the euRathlon 2015 challenge held in Piombino, Italy. Underwater conditions make it very difficult to use the RGB color space for object detection. Challenges include issues with underwater lighting, reflection and ray scattering. Initial testing was performed in a simulated environment using the UWSIM ROS package and live testing was primarily conducted at the competition on a Sparus II Autonomous Underwater Vehicle (AUV).