Open AccessThe extraction of maize lodgingregionsin UAV imagesusingdeepfully convolutional neural network
Author(s) -
Ergong Zheng,
Yanling Tian,
Zhaohong Xu,
Tao Chen
Publication year - 2020
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.179
H-Index - 26
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/474/3/032004
Subject(s) - convolutional neural network , extraction (chemistry) , computer science , artificial intelligence , pattern recognition (psychology) , remote sensing , artificial neural network , computer vision , geography , chemistry , chromatography
This paper presents a method for automatic extraction of maize lodging regions in unmanned aerial vehicle (UAV)visible images. A deep fully convolutional neural network is used to achieve the purpose.Firstly, images are collectedby using UAV remote sensing, and then the network is trained and validated on a total of more than 20000 labelled images, and finally, the network is able toaccurately extractlodging regions. Experimental results demonstrate the effectivenessof the method and F 1 score reaches 89.5% on test dataset. This study makes a major contribution to the assessment of maize lodging disaster.