Vision Aided Navigation of a Quad-Rotor for Autonomous Wind-Farm Inspection

Petar Durdevic; Daniel  Ortiz-Arroyo; Shaobao Li; Zhenyu Yang

doi:10.1016/j.ifacol.2019.08.049

Vision Aided Navigation of a Quad-Rotor for Autonomous Wind-Farm Inspection

Petar Durdevic, Daniel Ortiz-Arroyo, Shaobao Li, Zhenyu Yang

Research output: Contribution to journal › Conference article in Journal › Research › peer-review

13 Citations (Scopus)

432 Downloads (Pure)

Abstract

This work presents a vision based navigation system for an autonomous drone that is capable of recognizing and locating wind mills. WindMillNet, a Deep Neural Network created for this purpose was specially trained to recognize wind mills on camera images using transfer learning techniques. The drone powered by WindMillNet scans the horizon to find wind mills, and after perceiving a wind mill, navigates towards it, with the goal of performing its inspection. A hierarchical control system, implemented in the drone provides stability and control of its movements. Our framework was designed using a cyber-physical systems approach using high-level abstractions in modeling, communication, control and computation.

Original language	English
Book series	IFAC-PapersOnLine
Volume	52
Issue number	8
Pages (from-to)	61-66
Number of pages	6
ISSN	1474-6670
DOIs	https://doi.org/10.1016/j.ifacol.2019.08.049
Publication status	Published - Jul 2019
Event	IFAC Symposium Intelligent Autonomous Vehicles - 10th IAV 2019 - Gdansk, Poland Duration: 3 Jul 2019 → 5 Jul 2019 http://www.konsulting.gda.pl/iav2019/web/

Conference

Conference	IFAC Symposium Intelligent Autonomous Vehicles - 10th IAV 2019
Country/Territory	Poland
City	Gdansk
Period	03/07/2019 → 05/07/2019
Internet address	http://www.konsulting.gda.pl/iav2019/web/

Keywords

UAV
Control
Artificial Intelligence
Vision
Neural Networks
Deep Learning
Machine Learning

Access to Document

10.1016/j.ifacol.2019.08.049

1-s2.0-S2405896319303696-mainFinal published version, 1.07 MBLicence: CC BY-NC-ND 4.0

AUB Link

Search for the material in Aalborg University Library's search engine

Cite this

@inproceedings{168dffc1b8774790b6e229b0a494a305,

title = "Vision Aided Navigation of a Quad-Rotor for Autonomous Wind-Farm Inspection",

abstract = "This work presents a vision based navigation system for an autonomous drone that is capable of recognizing and locating wind mills. WindMillNet, a Deep Neural Network created for this purpose was specially trained to recognize wind mills on camera images using transfer learning techniques. The drone powered by WindMillNet scans the horizon to find wind mills, and after perceiving a wind mill, navigates towards it, with the goal of performing its inspection. A hierarchical control system, implemented in the drone provides stability and control of its movements. Our framework was designed using a cyber-physical systems approach using high-level abstractions in modeling, communication, control and computation.",

keywords = "UAV, Control, Artificial Intelligence, Vision, Neural Networks, Deep Learning, Machine Learning",

author = "Petar Durdevic and Daniel Ortiz-Arroyo and Shaobao Li and Zhenyu Yang",

year = "2019",

month = jul,

doi = "10.1016/j.ifacol.2019.08.049",

language = "English",

volume = "52",

pages = "61--66",

journal = "IFAC-PapersOnLine",

issn = "1474-6670",

publisher = "Elsevier",

number = "8",

note = "IFAC Symposium Intelligent Autonomous Vehicles - 10th IAV 2019 ; Conference date: 03-07-2019 Through 05-07-2019",

url = "http://www.konsulting.gda.pl/iav2019/web/",

}

TY - GEN

T1 - Vision Aided Navigation of a Quad-Rotor for Autonomous Wind-Farm Inspection

AU - Durdevic, Petar

AU - Ortiz-Arroyo, Daniel

AU - Li, Shaobao

AU - Yang, Zhenyu

PY - 2019/7

Y1 - 2019/7

N2 - This work presents a vision based navigation system for an autonomous drone that is capable of recognizing and locating wind mills. WindMillNet, a Deep Neural Network created for this purpose was specially trained to recognize wind mills on camera images using transfer learning techniques. The drone powered by WindMillNet scans the horizon to find wind mills, and after perceiving a wind mill, navigates towards it, with the goal of performing its inspection. A hierarchical control system, implemented in the drone provides stability and control of its movements. Our framework was designed using a cyber-physical systems approach using high-level abstractions in modeling, communication, control and computation.

AB - This work presents a vision based navigation system for an autonomous drone that is capable of recognizing and locating wind mills. WindMillNet, a Deep Neural Network created for this purpose was specially trained to recognize wind mills on camera images using transfer learning techniques. The drone powered by WindMillNet scans the horizon to find wind mills, and after perceiving a wind mill, navigates towards it, with the goal of performing its inspection. A hierarchical control system, implemented in the drone provides stability and control of its movements. Our framework was designed using a cyber-physical systems approach using high-level abstractions in modeling, communication, control and computation.

KW - UAV

KW - Control

KW - Artificial Intelligence

KW - Vision

KW - Neural Networks

KW - Deep Learning

KW - Machine Learning

U2 - 10.1016/j.ifacol.2019.08.049

DO - 10.1016/j.ifacol.2019.08.049

M3 - Conference article in Journal

SN - 1474-6670

VL - 52

SP - 61

EP - 66

JO - IFAC-PapersOnLine

JF - IFAC-PapersOnLine

IS - 8

T2 - IFAC Symposium Intelligent Autonomous Vehicles - 10th IAV 2019

Y2 - 3 July 2019 through 5 July 2019

ER -

Vision Aided Navigation of a Quad-Rotor for Autonomous Wind-Farm Inspection

Abstract

Conference

Keywords

Access to Document

AUB Link

Fingerprint

Cite this