Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment

This project aims to develop an autonomous UAV that is capable of performing higher-level visual task such as classification in real time in a hazy environment. Image haze removal is a challenging ill-posed problem and is a crucial image pre-processing step for the common Computer Vision (CV) tasks....

Full description

Saved in:
Bibliographic Details
Main Author: Chen, Jinkuan
Other Authors: Poenar Daniel Puiu
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2023
Subjects:
Online Access:https://hdl.handle.net/10356/167314
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-167314
record_format dspace
spelling sg-ntu-dr.10356-1673142023-07-07T15:47:38Z Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment Chen, Jinkuan Poenar Daniel Puiu School of Electrical and Electronic Engineering EPDPuiu@ntu.edu.sg Engineering::Electrical and electronic engineering This project aims to develop an autonomous UAV that is capable of performing higher-level visual task such as classification in real time in a hazy environment. Image haze removal is a challenging ill-posed problem and is a crucial image pre-processing step for the common Computer Vision (CV) tasks. Many approaches exist to remove noises caused by haze from the perspective of ground vehicles. However, there is limited research conducted on removing haze from the perspective of a UAV. This report presents the performance evaluations for seven existing dehazing approaches, based on either Convolutional Neural Networks (CNNs) or Vision Transformers (ViTs). These methods are evaluated using two existing datasets which serves as our benchmark, the AeroScapes and the UAVid, in which the hazy images are artificially generated via an established atmospheric model. This report also explored and evaluated the viability of using knowledge distillation on an existing dehazing model. The real time implementation of a dehazing procedure from a UAV is also documented. Bachelor of Engineering (Electrical and Electronic Engineering) 2023-05-25T08:07:47Z 2023-05-25T08:07:47Z 2023 Final Year Project (FYP) Chen, J. (2023). Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/167314 https://hdl.handle.net/10356/167314 en A2199-221 application/pdf Nanyang Technological University
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
spellingShingle Engineering::Electrical and electronic engineering
Chen, Jinkuan
Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment
description This project aims to develop an autonomous UAV that is capable of performing higher-level visual task such as classification in real time in a hazy environment. Image haze removal is a challenging ill-posed problem and is a crucial image pre-processing step for the common Computer Vision (CV) tasks. Many approaches exist to remove noises caused by haze from the perspective of ground vehicles. However, there is limited research conducted on removing haze from the perspective of a UAV. This report presents the performance evaluations for seven existing dehazing approaches, based on either Convolutional Neural Networks (CNNs) or Vision Transformers (ViTs). These methods are evaluated using two existing datasets which serves as our benchmark, the AeroScapes and the UAVid, in which the hazy images are artificially generated via an established atmospheric model. This report also explored and evaluated the viability of using knowledge distillation on an existing dehazing model. The real time implementation of a dehazing procedure from a UAV is also documented.
author2 Poenar Daniel Puiu
author_facet Poenar Daniel Puiu
Chen, Jinkuan
format Final Year Project
author Chen, Jinkuan
author_sort Chen, Jinkuan
title Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment
title_short Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment
title_full Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment
title_fullStr Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment
title_full_unstemmed Autonomous unmanned aerial vehicle (UAV) with computer-vision capabilities in a hazy environment
title_sort autonomous unmanned aerial vehicle (uav) with computer-vision capabilities in a hazy environment
publisher Nanyang Technological University
publishDate 2023
url https://hdl.handle.net/10356/167314
_version_ 1772825883807055872