North American Hardwoods Identification Using Machine-Learning

This technical note determines the feasibility of using an InceptionV4_ResNetV2 convolutional neural network (CNN) to correctly identify hardwood species from macroscopic images. The method is composed of a commodity smartphone fitted with a 14× macro lens for photography. The end-grains of...

Full description

Bibliographic Details
Main Authors:	Dercilio Junior Verly Lopes, Greg W. Burgreen, Edward D. Entsminger
Format:	Article
Language:	English
Published:	MDPI AG 2020-03-01
Series:	Forests
Subjects:	wood identification machine-learning smartphone macro lens inception-resnet convolutional neural networks (cnn)
Online Access:	https://www.mdpi.com/1999-4907/11/3/298

id	doaj-ad685398594e4d68b7174eb8fff12b57
record_format	Article
spelling	doaj-ad685398594e4d68b7174eb8fff12b572020-11-25T02:56:01ZengMDPI AGForests1999-49072020-03-0111329810.3390/f11030298f11030298North American Hardwoods Identification Using Machine-LearningDercilio Junior Verly Lopes0Greg W. Burgreen1Edward D. Entsminger2Department of Sustainable Bioproducts/Forest and Wildlife Research Center (FWRC), Mississippi State University, Starkville, MS 39762-9820, USACAVS, Mississippi State University, Starkville, MS 39759, USADepartment of Sustainable Bioproducts/Forest and Wildlife Research Center (FWRC), Mississippi State University, Starkville, MS 39762-9820, USAThis technical note determines the feasibility of using an InceptionV4_ResNetV2 convolutional neural network (CNN) to correctly identify hardwood species from macroscopic images. The method is composed of a commodity smartphone fitted with a 14× macro lens for photography. The end-grains of ten different North American hardwood species were photographed to create a dataset of 1869 images. The stratified 5-fold cross-validation machine-learning method was used, in which the number of testing samples varied from 341 to 342. Data augmentation was performed on-the-fly for each training set by rotating, zooming, and flipping images. It was found that the CNN could correctly identify hardwood species based on macroscopic images of its end-grain with an adjusted accuracy of 92.60%. With the current growing of machine-learning field, this model can then be readily deployed in a mobile application for field wood identification.https://www.mdpi.com/1999-4907/11/3/298wood identificationmachine-learningsmartphonemacro lensinception-resnetconvolutional neural networks (cnn)
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Dercilio Junior Verly Lopes Greg W. Burgreen Edward D. Entsminger
spellingShingle	Dercilio Junior Verly Lopes Greg W. Burgreen Edward D. Entsminger North American Hardwoods Identification Using Machine-Learning Forests wood identification machine-learning smartphone macro lens inception-resnet convolutional neural networks (cnn)
author_facet	Dercilio Junior Verly Lopes Greg W. Burgreen Edward D. Entsminger
author_sort	Dercilio Junior Verly Lopes
title	North American Hardwoods Identification Using Machine-Learning
title_short	North American Hardwoods Identification Using Machine-Learning
title_full	North American Hardwoods Identification Using Machine-Learning
title_fullStr	North American Hardwoods Identification Using Machine-Learning
title_full_unstemmed	North American Hardwoods Identification Using Machine-Learning
title_sort	north american hardwoods identification using machine-learning
publisher	MDPI AG
series	Forests
issn	1999-4907
publishDate	2020-03-01
description	This technical note determines the feasibility of using an InceptionV4_ResNetV2 convolutional neural network (CNN) to correctly identify hardwood species from macroscopic images. The method is composed of a commodity smartphone fitted with a 14× macro lens for photography. The end-grains of ten different North American hardwood species were photographed to create a dataset of 1869 images. The stratified 5-fold cross-validation machine-learning method was used, in which the number of testing samples varied from 341 to 342. Data augmentation was performed on-the-fly for each training set by rotating, zooming, and flipping images. It was found that the CNN could correctly identify hardwood species based on macroscopic images of its end-grain with an adjusted accuracy of 92.60%. With the current growing of machine-learning field, this model can then be readily deployed in a mobile application for field wood identification.
topic	wood identification machine-learning smartphone macro lens inception-resnet convolutional neural networks (cnn)
url	https://www.mdpi.com/1999-4907/11/3/298
work_keys_str_mv	AT derciliojuniorverlylopes northamericanhardwoodsidentificationusingmachinelearning AT gregwburgreen northamericanhardwoodsidentificationusingmachinelearning AT edwarddentsminger northamericanhardwoodsidentificationusingmachinelearning
_version_	1724714710633283584

North American Hardwoods Identification Using Machine-Learning

Similar Items