Design and Development of a UAV System for Multispectral Imaging and Remote Sensing Applications

• Author / Creator

  Marin Marcano, Jorge

• The increasing use of the Unmanned Aerial Vehicles (UAVs) over the past few years has opened the door to a large number of applications in research, commercial, and industrial areas. UAVs have become an efficient, cost effective, and environmentally friendly (less or zero CO2 emissions) platform for remote sensing compared to manned vehicles. One major field in this category is multispectral sensing. Commercial solutions of UAV-compatible multispectral cameras are available but tend to have a high cost and, in most cases, lack certain flexibility when it comes to custom development. This thesis is organized in three basic sections. It begins with the development of a modular fixed-wing UAV that includes a customized airframe, integrated avionics with a custom UAV system board with redundant power supply, and single board computer (SBC) for payload interfacing. The other two sections of this thesis are focused on the design, development, and testing of two revisions of a custom multispectral imaging system. The customized fixed-wing airframe was redesigned aiming for modularity and reliability. The open-source UAV system board integrates all essential sensors and components in a small and compact board, reducing the amount of wiring and connectors, which in turn reduces weight and increases reliability. The UAV system board includes a detachable and novel redundant power supply board that provides main and backup power regulation. The complete UAV system design is presented as an efficient, and flexible platform. The multispectral imaging system is based on multiple USB 2.0 camera modules and optical filters. The first iteration, or proof of concept, introduces a 3-band multispectral imaging system based on 3 rolling shutter cameras synchronized by software, with optical filters for the green, red, and near-infrared bands. The cameras were controlled by the SBC on board of the UAV. The imaging system was tested over a grass field, and the images collected were post-processed. Resulting images were aligned and NDVI images were produced where it was possible to detect vegetation and identify different levels of vegetation stress. This system proved to be a low cost and simple, yet effective approach for custom UAV multispectral imaging. The next evolution of the system is a 12-band multispectral imaging platform based on USB 2.0 camera modules. Cameras used in this revision have a global shutter and a trigger functioning mode that allows hardware synchronization. A novel bridge board was designed to integrate all USB communications via USB HUBs and also include a microcontroller unit (MCU) in charge of triggering the cameras. The USB HUB network has selectable routes for grouping cameras to different upstream USB ports depending on the bandwidth requirements of the application. The system was tested and characterized for the maximum number of simultaneous cameras per root USB port depending on camera configuration. With the highly integrated bridge board the overall size and weight of the system was reduced, making it compatible with most UAVs. Software was also developed, based on third party tools, that allows extensive configuration of the cameras and control over the imaging system. The result of this work introduces a detailed methodology for UAV systems development, introduces a comprehensive UAV system board, and explores the capacities of USB 2.0 camera modules as part of complex imaging systems. The hardware and software platforms developed throughout this thesis are cost effective and efficient solutions that allow customization and expansion, making them an ideal development tool for UAV systems and UAV multispectral imaging applications.

• Subjects / Keywords

  • USB Cameras
  • Flight Controller
  • Unmanned Aerial Vehicle (UAV)
  • Multispectral Imaging
  • Remote Sensing

• Graduation date

  Spring 2018

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3JW8731J

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.