Cooperative robotic exploration

• Main Author: Foo, Sean Yuan Zhang
• Other Authors: Sundaram Suresh
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/70324
• Institution: Nanyang Technological University

Drones are originally defined as Unmanned Aerial Vehicles or UAVs that operate under varying degrees of autonomy. The definition of drone has however come to cover Non-Aerial vehicles as well. Drones have a plethora of applications that cover industrial, commercial, military and consumer related fields. The recent advent of the easily accessible consumer drone, has paved the way for companies like Parrot to create Jumping Night and Jumping Race drones for consumer use. One aspect of drone autonomy has been to use computer vision techniques to enable drones to gain a level of understanding of visual data via camera inputs. Computer vision techniques provide ways to process and analyse visual information that the drone can use to recognise objects and obstacles within the environment. The project aims to create a multiple robotic system for the purpose of cooperative exploration within a space. Using both the Jumping Night and Jumping Race drones, they would identify a known object within the space and follow it. Using the Computer Vision Library OpenCV, they would be able to identify a shape and colour, specified within the algorithm and accurately track the object in question. This report aims to detail the development of this system. It will include the documentation of the equipment and technology used and the actual development process. It is the goal of this system to provide an accessible means of exploration via the use of such consumer drones as well as to be a guide for the further development of these drones in this context.