Deep learning for use in lumber classification tasks
Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learni...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Hu, Junfeng [verfasserIn] |
|---|
| Format: |
Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2019 |
|---|
| Anmerkung: |
© Springer-Verlag GmbH Germany, part of Springer Nature 2019 |
|---|
| Übergeordnetes Werk: |
Enthalten in: Wood science and technology - Springer Berlin Heidelberg, 1967, 53(2019), 2 vom: 26. Feb., Seite 505-517 |
|---|---|
| Übergeordnetes Werk: |
volume:53 ; year:2019 ; number:2 ; day:26 ; month:02 ; pages:505-517 |
| Links: |
|---|
| DOI / URN: |
10.1007/s00226-019-01086-z |
|---|
| Katalog-ID: |
OLC2073079393 |
|---|
| LEADER | 01000caa a22002652 4500 | ||
|---|---|---|---|
| 001 | OLC2073079393 | ||
| 003 | DE-627 | ||
| 005 | 20230323230730.0 | ||
| 007 | tu | ||
| 008 | 200819s2019 xx ||||| 00| ||eng c | ||
| 024 | 7 | |a 10.1007/s00226-019-01086-z |2 doi | |
| 035 | |a (DE-627)OLC2073079393 | ||
| 035 | |a (DE-He213)s00226-019-01086-z-p | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 670 |q VZ |
| 084 | |a 23 |2 ssgn | ||
| 100 | 1 | |a Hu, Junfeng |e verfasserin |0 (orcid)0000-0002-1174-385X |4 aut | |
| 245 | 1 | 0 | |a Deep learning for use in lumber classification tasks |
| 264 | 1 | |c 2019 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ohne Hilfsmittel zu benutzen |b n |2 rdamedia | ||
| 338 | |a Band |b nc |2 rdacarrier | ||
| 500 | |a © Springer-Verlag GmbH Germany, part of Springer Nature 2019 | ||
| 520 | |a Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. | ||
| 700 | 1 | |a Song, Wenlong |4 aut | |
| 700 | 1 | |a Zhang, Wei |4 aut | |
| 700 | 1 | |a Zhao, Yafeng |4 aut | |
| 700 | 1 | |a Yilmaz, Alper |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Wood science and technology |d Springer Berlin Heidelberg, 1967 |g 53(2019), 2 vom: 26. Feb., Seite 505-517 |w (DE-627)129600679 |w (DE-600)241313-9 |w (DE-576)015094227 |x 0043-7719 |7 nnns |
| 773 | 1 | 8 | |g volume:53 |g year:2019 |g number:2 |g day:26 |g month:02 |g pages:505-517 |
| 856 | 4 | 1 | |u https://doi.org/10.1007/s00226-019-01086-z |z lizenzpflichtig |3 Volltext |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_OLC | ||
| 912 | |a SSG-OLC-TEC | ||
| 912 | |a SSG-OLC-CHE | ||
| 912 | |a SSG-OLC-FOR | ||
| 912 | |a SSG-OPC-FOR | ||
| 912 | |a GBV_ILN_70 | ||
| 912 | |a GBV_ILN_2016 | ||
| 912 | |a GBV_ILN_2018 | ||
| 912 | |a GBV_ILN_2542 | ||
| 912 | |a GBV_ILN_4277 | ||
| 951 | |a AR | ||
| 952 | |d 53 |j 2019 |e 2 |b 26 |c 02 |h 505-517 | ||
| author_variant |
j h jh w s ws w z wz y z yz a y ay |
|---|---|
| matchkey_str |
article:00437719:2019----::eperigouenubrlsi |
| hierarchy_sort_str |
2019 |
| publishDate |
2019 |
| allfields |
10.1007/s00226-019-01086-z doi (DE-627)OLC2073079393 (DE-He213)s00226-019-01086-z-p DE-627 ger DE-627 rakwb eng 670 VZ 23 ssgn Hu, Junfeng verfasserin (orcid)0000-0002-1174-385X aut Deep learning for use in lumber classification tasks 2019 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. Song, Wenlong aut Zhang, Wei aut Zhao, Yafeng aut Yilmaz, Alper aut Enthalten in Wood science and technology Springer Berlin Heidelberg, 1967 53(2019), 2 vom: 26. Feb., Seite 505-517 (DE-627)129600679 (DE-600)241313-9 (DE-576)015094227 0043-7719 nnns volume:53 year:2019 number:2 day:26 month:02 pages:505-517 https://doi.org/10.1007/s00226-019-01086-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-CHE SSG-OLC-FOR SSG-OPC-FOR GBV_ILN_70 GBV_ILN_2016 GBV_ILN_2018 GBV_ILN_2542 GBV_ILN_4277 AR 53 2019 2 26 02 505-517 |
| spelling |
10.1007/s00226-019-01086-z doi (DE-627)OLC2073079393 (DE-He213)s00226-019-01086-z-p DE-627 ger DE-627 rakwb eng 670 VZ 23 ssgn Hu, Junfeng verfasserin (orcid)0000-0002-1174-385X aut Deep learning for use in lumber classification tasks 2019 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. Song, Wenlong aut Zhang, Wei aut Zhao, Yafeng aut Yilmaz, Alper aut Enthalten in Wood science and technology Springer Berlin Heidelberg, 1967 53(2019), 2 vom: 26. Feb., Seite 505-517 (DE-627)129600679 (DE-600)241313-9 (DE-576)015094227 0043-7719 nnns volume:53 year:2019 number:2 day:26 month:02 pages:505-517 https://doi.org/10.1007/s00226-019-01086-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-CHE SSG-OLC-FOR SSG-OPC-FOR GBV_ILN_70 GBV_ILN_2016 GBV_ILN_2018 GBV_ILN_2542 GBV_ILN_4277 AR 53 2019 2 26 02 505-517 |
| allfields_unstemmed |
10.1007/s00226-019-01086-z doi (DE-627)OLC2073079393 (DE-He213)s00226-019-01086-z-p DE-627 ger DE-627 rakwb eng 670 VZ 23 ssgn Hu, Junfeng verfasserin (orcid)0000-0002-1174-385X aut Deep learning for use in lumber classification tasks 2019 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. Song, Wenlong aut Zhang, Wei aut Zhao, Yafeng aut Yilmaz, Alper aut Enthalten in Wood science and technology Springer Berlin Heidelberg, 1967 53(2019), 2 vom: 26. Feb., Seite 505-517 (DE-627)129600679 (DE-600)241313-9 (DE-576)015094227 0043-7719 nnns volume:53 year:2019 number:2 day:26 month:02 pages:505-517 https://doi.org/10.1007/s00226-019-01086-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-CHE SSG-OLC-FOR SSG-OPC-FOR GBV_ILN_70 GBV_ILN_2016 GBV_ILN_2018 GBV_ILN_2542 GBV_ILN_4277 AR 53 2019 2 26 02 505-517 |
| allfieldsGer |
10.1007/s00226-019-01086-z doi (DE-627)OLC2073079393 (DE-He213)s00226-019-01086-z-p DE-627 ger DE-627 rakwb eng 670 VZ 23 ssgn Hu, Junfeng verfasserin (orcid)0000-0002-1174-385X aut Deep learning for use in lumber classification tasks 2019 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. Song, Wenlong aut Zhang, Wei aut Zhao, Yafeng aut Yilmaz, Alper aut Enthalten in Wood science and technology Springer Berlin Heidelberg, 1967 53(2019), 2 vom: 26. Feb., Seite 505-517 (DE-627)129600679 (DE-600)241313-9 (DE-576)015094227 0043-7719 nnns volume:53 year:2019 number:2 day:26 month:02 pages:505-517 https://doi.org/10.1007/s00226-019-01086-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-CHE SSG-OLC-FOR SSG-OPC-FOR GBV_ILN_70 GBV_ILN_2016 GBV_ILN_2018 GBV_ILN_2542 GBV_ILN_4277 AR 53 2019 2 26 02 505-517 |
| allfieldsSound |
10.1007/s00226-019-01086-z doi (DE-627)OLC2073079393 (DE-He213)s00226-019-01086-z-p DE-627 ger DE-627 rakwb eng 670 VZ 23 ssgn Hu, Junfeng verfasserin (orcid)0000-0002-1174-385X aut Deep learning for use in lumber classification tasks 2019 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier © Springer-Verlag GmbH Germany, part of Springer Nature 2019 Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. Song, Wenlong aut Zhang, Wei aut Zhao, Yafeng aut Yilmaz, Alper aut Enthalten in Wood science and technology Springer Berlin Heidelberg, 1967 53(2019), 2 vom: 26. Feb., Seite 505-517 (DE-627)129600679 (DE-600)241313-9 (DE-576)015094227 0043-7719 nnns volume:53 year:2019 number:2 day:26 month:02 pages:505-517 https://doi.org/10.1007/s00226-019-01086-z lizenzpflichtig Volltext GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-CHE SSG-OLC-FOR SSG-OPC-FOR GBV_ILN_70 GBV_ILN_2016 GBV_ILN_2018 GBV_ILN_2542 GBV_ILN_4277 AR 53 2019 2 26 02 505-517 |
| language |
English |
| source |
Enthalten in Wood science and technology 53(2019), 2 vom: 26. Feb., Seite 505-517 volume:53 year:2019 number:2 day:26 month:02 pages:505-517 |
| sourceStr |
Enthalten in Wood science and technology 53(2019), 2 vom: 26. Feb., Seite 505-517 volume:53 year:2019 number:2 day:26 month:02 pages:505-517 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| dewey-raw |
670 |
| isfreeaccess_bool |
false |
| container_title |
Wood science and technology |
| authorswithroles_txt_mv |
Hu, Junfeng @@aut@@ Song, Wenlong @@aut@@ Zhang, Wei @@aut@@ Zhao, Yafeng @@aut@@ Yilmaz, Alper @@aut@@ |
| publishDateDaySort_date |
2019-02-26T00:00:00Z |
| hierarchy_top_id |
129600679 |
| dewey-sort |
3670 |
| id |
OLC2073079393 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">OLC2073079393</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230323230730.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">200819s2019 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s00226-019-01086-z</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2073079393</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s00226-019-01086-z-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">670</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">23</subfield><subfield code="2">ssgn</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Hu, Junfeng</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0002-1174-385X</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Deep learning for use in lumber classification tasks</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2019</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© Springer-Verlag GmbH Germany, part of Springer Nature 2019</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU.</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Song, Wenlong</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhang, Wei</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhao, Yafeng</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Yilmaz, Alper</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Wood science and technology</subfield><subfield code="d">Springer Berlin Heidelberg, 1967</subfield><subfield code="g">53(2019), 2 vom: 26. Feb., Seite 505-517</subfield><subfield code="w">(DE-627)129600679</subfield><subfield code="w">(DE-600)241313-9</subfield><subfield code="w">(DE-576)015094227</subfield><subfield code="x">0043-7719</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:53</subfield><subfield code="g">year:2019</subfield><subfield code="g">number:2</subfield><subfield code="g">day:26</subfield><subfield code="g">month:02</subfield><subfield code="g">pages:505-517</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s00226-019-01086-z</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-TEC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-CHE</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-FOR</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OPC-FOR</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2016</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2018</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2542</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4277</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">53</subfield><subfield code="j">2019</subfield><subfield code="e">2</subfield><subfield code="b">26</subfield><subfield code="c">02</subfield><subfield code="h">505-517</subfield></datafield></record></collection>
|
| author |
Hu, Junfeng |
| spellingShingle |
Hu, Junfeng ddc 670 ssgn 23 Deep learning for use in lumber classification tasks |
| authorStr |
Hu, Junfeng |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)129600679 |
| format |
Article |
| dewey-ones |
670 - Manufacturing |
| delete_txt_mv |
keep |
| author_role |
aut aut aut aut aut |
| collection |
OLC |
| remote_str |
false |
| illustrated |
Not Illustrated |
| issn |
0043-7719 |
| topic_title |
670 VZ 23 ssgn Deep learning for use in lumber classification tasks |
| topic |
ddc 670 ssgn 23 |
| topic_unstemmed |
ddc 670 ssgn 23 |
| topic_browse |
ddc 670 ssgn 23 |
| format_facet |
Aufsätze Gedruckte Aufsätze |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
nc |
| hierarchy_parent_title |
Wood science and technology |
| hierarchy_parent_id |
129600679 |
| dewey-tens |
670 - Manufacturing |
| hierarchy_top_title |
Wood science and technology |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)129600679 (DE-600)241313-9 (DE-576)015094227 |
| title |
Deep learning for use in lumber classification tasks |
| ctrlnum |
(DE-627)OLC2073079393 (DE-He213)s00226-019-01086-z-p |
| title_full |
Deep learning for use in lumber classification tasks |
| author_sort |
Hu, Junfeng |
| journal |
Wood science and technology |
| journalStr |
Wood science and technology |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
600 - Technology |
| recordtype |
marc |
| publishDateSort |
2019 |
| contenttype_str_mv |
txt |
| container_start_page |
505 |
| author_browse |
Hu, Junfeng Song, Wenlong Zhang, Wei Zhao, Yafeng Yilmaz, Alper |
| container_volume |
53 |
| class |
670 VZ 23 ssgn |
| format_se |
Aufsätze |
| author-letter |
Hu, Junfeng |
| doi_str_mv |
10.1007/s00226-019-01086-z |
| normlink |
(ORCID)0000-0002-1174-385X |
| normlink_prefix_str_mv |
(orcid)0000-0002-1174-385X |
| dewey-full |
670 |
| title_sort |
deep learning for use in lumber classification tasks |
| title_auth |
Deep learning for use in lumber classification tasks |
| abstract |
Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. © Springer-Verlag GmbH Germany, part of Springer Nature 2019 |
| abstractGer |
Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. © Springer-Verlag GmbH Germany, part of Springer Nature 2019 |
| abstract_unstemmed |
Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU. © Springer-Verlag GmbH Germany, part of Springer Nature 2019 |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-CHE SSG-OLC-FOR SSG-OPC-FOR GBV_ILN_70 GBV_ILN_2016 GBV_ILN_2018 GBV_ILN_2542 GBV_ILN_4277 |
| container_issue |
2 |
| title_short |
Deep learning for use in lumber classification tasks |
| url |
https://doi.org/10.1007/s00226-019-01086-z |
| remote_bool |
false |
| author2 |
Song, Wenlong Zhang, Wei Zhao, Yafeng Yilmaz, Alper |
| author2Str |
Song, Wenlong Zhang, Wei Zhao, Yafeng Yilmaz, Alper |
| ppnlink |
129600679 |
| mediatype_str_mv |
n |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| doi_str |
10.1007/s00226-019-01086-z |
| up_date |
2024-07-03T17:11:51.299Z |
| _version_ |
1803578728831254528 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a22002652 4500</leader><controlfield tag="001">OLC2073079393</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20230323230730.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">200819s2019 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1007/s00226-019-01086-z</subfield><subfield code="2">doi</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC2073079393</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-He213)s00226-019-01086-z-p</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">670</subfield><subfield code="q">VZ</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">23</subfield><subfield code="2">ssgn</subfield></datafield><datafield tag="100" ind1="1" ind2=" "><subfield code="a">Hu, Junfeng</subfield><subfield code="e">verfasserin</subfield><subfield code="0">(orcid)0000-0002-1174-385X</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Deep learning for use in lumber classification tasks</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2019</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="500" ind1=" " ind2=" "><subfield code="a">© Springer-Verlag GmbH Germany, part of Springer Nature 2019</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Abstract Digital image processing has been widely used in the wood industry, and it is expected that the algorithms are accurate and have low computational complexities, especially for the real-time lumber grading and lumber classification processes. This paper investigates variations of deep learning strategies based on ResNet18 for classification of lumber images. The four datasets used in this work were manually marked as lumber defects, wood textures and lumbers by experts. A key ideal is to employ the transfer learning in the context of convolutional neural networks with a classifier layer only training with a small amount of training data for different tasks at the same lumber machinery. Through the expansion of unbalanced samples, the accuracy rate has been effectively improved. The human involvement when needed is kept to a minimum only for the training phase. The proposed approach was independently tested with four datasets, of which 80% of the data is used for training and 20% of the data is used for testing. The classification accuracy of the approach for each of the datasets is 98.16%, 93.32%, 96.64% and 99.50%. The average time for sorting the lumber image was kept at 0.003 s when the system runs on Nvidia GTX860 GPU.</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Song, Wenlong</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhang, Wei</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Zhao, Yafeng</subfield><subfield code="4">aut</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Yilmaz, Alper</subfield><subfield code="4">aut</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">Wood science and technology</subfield><subfield code="d">Springer Berlin Heidelberg, 1967</subfield><subfield code="g">53(2019), 2 vom: 26. Feb., Seite 505-517</subfield><subfield code="w">(DE-627)129600679</subfield><subfield code="w">(DE-600)241313-9</subfield><subfield code="w">(DE-576)015094227</subfield><subfield code="x">0043-7719</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:53</subfield><subfield code="g">year:2019</subfield><subfield code="g">number:2</subfield><subfield code="g">day:26</subfield><subfield code="g">month:02</subfield><subfield code="g">pages:505-517</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">https://doi.org/10.1007/s00226-019-01086-z</subfield><subfield code="z">lizenzpflichtig</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-TEC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-CHE</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-FOR</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OPC-FOR</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2016</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2018</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2542</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_4277</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">53</subfield><subfield code="j">2019</subfield><subfield code="e">2</subfield><subfield code="b">26</subfield><subfield code="c">02</subfield><subfield code="h">505-517</subfield></datafield></record></collection>
|
| score |
7.3995953 |