Please use this identifier to cite or link to this item:
Title: Power Line Recognition From Aerial Images With Deep Learning
Authors: Yetgin, Ömer Emre
Benligiray, Burak
Gerek, Ömer Nezih
Keywords: Feature extraction
Task analysis
Image edge detection
Convolutional neural networks
Collision avoidance
deep learning
flight assistance
power line recognition
Issue Date: 2019
Publisher: IEEE-Inst Electrical Electronics Engineers Inc
Abstract: Avoidance of power lines is an important issue of flight safety. Assistance systems that automatically detect power lines can prevent accidents caused by pilot unawareness. In this study, we propose using convolutional neural networks (CNN) to recognize the presence of power lines in aerial images. Deep CNN architectures such as VGG and ResNet are originally designed to recognize objects in the ImageNet dataset. We show that they are also successful at extracting features that indicate the presence of power lines, which appear as simple, yet subtle structures. Another interesting finding is that pretraining the CNN with the ImageNet dataset improves power line recognition rate significantly. This indicates that the usage of ImageNet pretraining should not be limited to high-level visual tasks, as it also develops general-purpose visual skills that apply to more primitive tasks. To test the proposed methods' performance, we collected an aerial dataset and made it publicly available. We experimented with training CNNs in an end-to-end fashion, along with extracting features from the intermediate stages of CNNs and feeding them to various classifiers. These experiments were repeated with different architectures and preprocessing methods, resulting in an expansive account of best practices for the usage of CNNs for power line recognition.
ISSN: 0018-9251
Appears in Collections:Elektrik-Elektronik Mühendisliği Bölümü Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu
WoS İndeksli Yayınlar Koleksiyonu

Show full item record

CORE Recommender


checked on Jun 22, 2022

Page view(s)

checked on Oct 3, 2022

Google ScholarTM



Items in GCRIS Repository are protected by copyright, with all rights reserved, unless otherwise indicated.