Design of multi-scale receptive field convolutional neural network for surface inspection of hot rolled steels
Due to the large intra-class variations and unbalanced training samples, the accuracy of existing algorithms used in defect classification of hot rolled steels is unsatisfactory. In this paper, a new hierarchical learning framework is proposed based on convolutional neural networks to classify hot r...
Uloženo v:
| Vydáno v: | Image and vision computing Ročník 89; s. 12 - 20 |
|---|---|
| Hlavní autoři: | , , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
Elsevier B.V
01.09.2019
|
| Témata: | |
| ISSN: | 0262-8856, 1872-8138 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | Due to the large intra-class variations and unbalanced training samples, the accuracy of existing algorithms used in defect classification of hot rolled steels is unsatisfactory. In this paper, a new hierarchical learning framework is proposed based on convolutional neural networks to classify hot rolled defects. Multi-scale receptive field is introduced in the new framework to extract multi-scale features, which can better represent defects than the feature maps produced by a single convolutional layer. A group of AutoEncoders are trained to reduce the dimension of the extracted multi-scale features which improve the generalization ability under insufficient training samples. Besides, to mitigate the deviation caused by fine-tuning the pre-trained model with images of different context, we add a penalty term in the loss function, which is to reconstruct the input image from the feature maps produced by the pre-trained model, to help network encode more effective and structured information. The experiments with samples captured from two hot rolled production lines showed that the proposed framework achieved a classification rate of 97.2% and 97% respectively, which are much higher than the conventional methods.
•We proposed a new framework based on Inception-V4 integrated with multi-scale respective field.•AutoEncoders are trained to reduce the dimension of the multi-scale features extracted by the pre-trained model.•A penalty term in the loss function to reconstruct the input image from the feature maps was added.•The methods were tested with two different samples from online surface inspection system of steels. |
|---|---|
| AbstractList | Due to the large intra-class variations and unbalanced training samples, the accuracy of existing algorithms used in defect classification of hot rolled steels is unsatisfactory. In this paper, a new hierarchical learning framework is proposed based on convolutional neural networks to classify hot rolled defects. Multi-scale receptive field is introduced in the new framework to extract multi-scale features, which can better represent defects than the feature maps produced by a single convolutional layer. A group of AutoEncoders are trained to reduce the dimension of the extracted multi-scale features which improve the generalization ability under insufficient training samples. Besides, to mitigate the deviation caused by fine-tuning the pre-trained model with images of different context, we add a penalty term in the loss function, which is to reconstruct the input image from the feature maps produced by the pre-trained model, to help network encode more effective and structured information. The experiments with samples captured from two hot rolled production lines showed that the proposed framework achieved a classification rate of 97.2% and 97% respectively, which are much higher than the conventional methods.
•We proposed a new framework based on Inception-V4 integrated with multi-scale respective field.•AutoEncoders are trained to reduce the dimension of the multi-scale features extracted by the pre-trained model.•A penalty term in the loss function to reconstruct the input image from the feature maps was added.•The methods were tested with two different samples from online surface inspection system of steels. |
| Author | Wang, Dadong Xu, Ke He, Di |
| Author_xml | – sequence: 1 givenname: Di surname: He fullname: He, Di organization: Collaborative Innovation Center of Steel Technology, University of Science and Technology Beijing, Beijing 100083, China – sequence: 2 givenname: Ke surname: Xu fullname: Xu, Ke email: xuke@ustb.edu.cn organization: Collaborative Innovation Center of Steel Technology, University of Science and Technology Beijing, Beijing 100083, China – sequence: 3 givenname: Dadong surname: Wang fullname: Wang, Dadong organization: Quantitative Imaging Research Team, Data61, Commonwealth Scientific and Industrial Research Organization, Australia |
| BookMark | eNqFkMtOwzAQRS1UJNrCH7DwDyQ4iWs7LJBQeUqV2MDacp0xuLh2ZTtF_D0JZcUCFqM7izlXmjNDEx88IHRekbIiFbvYlHar9jaVNanakrCSEHGEppXgdSGqRkzQlNRs2MWCnaBZShtCCCe8nSJ_A8m-ehwM3vYu2yJp5QBH0LDLdg_YWHAd1sHvg-uzDV457KGP35E_QnzHJkSc-miUBmx92oEe78bKt5BxDM5Bh1MGcOkUHRvlEpz95By93N0-Lx-K1dP94_J6VWhaiVy0VNDWrIfRCijUHWeGmIWm3KxNC6auBVdKUNo0a80YCM4bQzrV8KZrCdBmjuihV8eQUgQjd3FwFD9lReToTG7kwZkcnUnC5OBswC5_YdpmNX6To7LuP_jqAA9_wt5ClElb8Bo6O-jMsgv274Iv8KuQPQ |
| CitedBy_id | crossref_primary_10_3390_coatings13010017 crossref_primary_10_1109_TIM_2020_3030167 crossref_primary_10_1007_s13369_025_10223_9 crossref_primary_10_1016_j_rcim_2020_102083 crossref_primary_10_32604_jai_2022_038875 crossref_primary_10_3390_met11111832 crossref_primary_10_1109_TCPMT_2020_3033837 crossref_primary_10_3390_jimaging10060138 crossref_primary_10_1117_1_JEI_31_6_063056 crossref_primary_10_1049_ipr2_12647 crossref_primary_10_1080_0951192X_2021_1901319 crossref_primary_10_1109_TIM_2021_3063755 crossref_primary_10_3390_s21134612 crossref_primary_10_3390_pr9010033 crossref_primary_10_1080_00207543_2022_2138613 crossref_primary_10_1016_j_image_2022_116807 crossref_primary_10_1109_JSEN_2021_3131645 crossref_primary_10_1109_ACCESS_2024_3447035 crossref_primary_10_1016_j_jprocont_2022_06_014 crossref_primary_10_1109_TASE_2022_3140784 crossref_primary_10_3390_app11209473 crossref_primary_10_1016_j_ymeth_2020_05_012 crossref_primary_10_3390_s21144870 crossref_primary_10_3390_informatics11020025 crossref_primary_10_1109_JSEN_2023_3263924 crossref_primary_10_1016_j_matlet_2021_129707 crossref_primary_10_1155_2021_6637252 crossref_primary_10_1002_srin_202200505 |
| Cites_doi | 10.1117/1.OE.51.11.113605 10.1016/j.imavis.2015.01.001 10.3390/met7080311 10.1109/CVPR.2017.690 10.1016/j.ndteint.2009.01.007 10.1109/5.726791 10.1016/S1006-706X(13)60102-8 |
| ContentType | Journal Article |
| Copyright | 2019 Elsevier B.V. |
| Copyright_xml | – notice: 2019 Elsevier B.V. |
| DBID | AAYXX CITATION |
| DOI | 10.1016/j.imavis.2019.06.008 |
| DatabaseName | CrossRef |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Applied Sciences Engineering |
| EISSN | 1872-8138 |
| EndPage | 20 |
| ExternalDocumentID | 10_1016_j_imavis_2019_06_008 S0262885619300915 |
| GroupedDBID | --K --M .~1 0R~ 1B1 1~. 1~5 29I 4.4 457 4G. 5GY 5VS 7-5 71M 8P~ 9JN AABNK AACTN AAEDT AAEDW AAIAV AAIKJ AAKOC AALRI AAOAW AAQFI AAQXK AAXUO AAYFN ABBOA ABFNM ABFRF ABJNI ABMAC ABOCM ABTAH ABXDB ABYKQ ACDAQ ACGFO ACGFS ACNNM ACRLP ACZNC ADBBV ADEZE ADJOM ADMUD ADTZH AEBSH AECPX AEFWE AEKER AENEX AFKWA AFTJW AGHFR AGUBO AGYEJ AHHHB AHJVU AHZHX AIALX AIEXJ AIKHN AITUG AJBFU AJOXV ALMA_UNASSIGNED_HOLDINGS AMFUW AMRAJ AOUOD ASPBG AVWKF AXJTR AZFZN BJAXD BKOJK BLXMC CS3 DU5 EBS EFJIC EFLBG EJD EO8 EO9 EP2 EP3 F0J F5P FDB FEDTE FGOYB FIRID FNPLU FYGXN G-Q G8K GBLVA GBOLZ HLZ HVGLF HZ~ IHE J1W JJJVA KOM LG9 M41 MO0 N9A O-L O9- OAUVE OZT P-8 P-9 P2P PC. Q38 R2- RIG RNS ROL RPZ SBC SDF SDG SDP SES SEW SPC SPCBC SST SSV SSZ T5K TN5 UHS UNMZH VOH WUQ XFK XPP ZMT ZY4 ~G- 9DU AATTM AAXKI AAYWO AAYXX ABDPE ABWVN ACLOT ACRPL ACVFH ADCNI ADNMO AEIPS AEUPX AFJKZ AFPUW AGQPQ AIGII AIIUN AKBMS AKRWK AKYEP ANKPU APXCP CITATION EFKBS ~HD |
| ID | FETCH-LOGICAL-c418t-94849fb49fcae4e2d76f0f5c47fbf9ef2287aa84433bc66e8773f0da373d90e43 |
| ISICitedReferencesCount | 34 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000487574900002&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0262-8856 |
| IngestDate | Sat Nov 29 07:21:26 EST 2025 Tue Nov 18 21:53:46 EST 2025 Fri Feb 23 02:23:40 EST 2024 |
| IsDoiOpenAccess | false |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Defect identification Surface inspection Hot rolled steels Convolutional neural networks AutoEncoder |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c418t-94849fb49fcae4e2d76f0f5c47fbf9ef2287aa84433bc66e8773f0da373d90e43 |
| OpenAccessLink | http://hdl.handle.net/1959.4/unsworks_75624 |
| PageCount | 9 |
| ParticipantIDs | crossref_primary_10_1016_j_imavis_2019_06_008 crossref_citationtrail_10_1016_j_imavis_2019_06_008 elsevier_sciencedirect_doi_10_1016_j_imavis_2019_06_008 |
| PublicationCentury | 2000 |
| PublicationDate | September 2019 2019-09-00 |
| PublicationDateYYYYMMDD | 2019-09-01 |
| PublicationDate_xml | – month: 09 year: 2019 text: September 2019 |
| PublicationDecade | 2010 |
| PublicationTitle | Image and vision computing |
| PublicationYear | 2019 |
| Publisher | Elsevier B.V |
| Publisher_xml | – name: Elsevier B.V |
| References | Szegedy (bb0060) 2015 Szegedy (bb0070) 2017 Van Hulle (bb0030) 2012 Tian, Xu (bb0015) 2017; 7 Redmon (bb0090) 2016 AI Yong-hao (bb0115) 2013; 20 Krizhevsky, Sutskever, Hinton (bb0045) 2012 Redmon, Joseph, Ali Farhadi, 2016. YOLO9000: better, faster, stronger. arXiv preprint Deng (bb0105) 2013 Paulraj, Shukry, Yaacob (bb0020) 2010 Xu, Liu, Ai (bb0010) 2015; 35 Kaiming (bb0065) 2016 Lin M., Chen Q., Yan S., 2013. Network in network. arXiv preprinted Simonyan K., Zisserman A., 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint Ai, Xu (bb0110) 2012; 51 Ren (bb0085) 2015 . Girshick (bb0080) 2015 Karpathy (bb0100) 2014 Xu, Xu, Chen (bb0005) 2002; 3 Yun, Choi, Kim (bb0025) 2009; 42 LeCun, Bottou, Bengio (bb0040) 1998; 86 Girshick (bb0075) 2014 Suvdaa, Ahn, Ko (bb0035) 2012; 6 Xu, Xu, Lu (bb0120) 1999; 6 10.1016/j.imavis.2019.06.008_bb0095 10.1016/j.imavis.2019.06.008_bb0050 Deng (10.1016/j.imavis.2019.06.008_bb0105) 2013 Ai (10.1016/j.imavis.2019.06.008_bb0110) 2012; 51 Van Hulle (10.1016/j.imavis.2019.06.008_bb0030) 2012 Kaiming (10.1016/j.imavis.2019.06.008_bb0065) 2016 Szegedy (10.1016/j.imavis.2019.06.008_bb0070) 2017 Girshick (10.1016/j.imavis.2019.06.008_bb0075) 2014 Yun (10.1016/j.imavis.2019.06.008_bb0025) 2009; 42 Girshick (10.1016/j.imavis.2019.06.008_bb0080) 2015 Szegedy (10.1016/j.imavis.2019.06.008_bb0060) 2015 Xu (10.1016/j.imavis.2019.06.008_bb0010) 2015; 35 Karpathy (10.1016/j.imavis.2019.06.008_bb0100) 2014 Redmon (10.1016/j.imavis.2019.06.008_bb0090) 2016 Suvdaa (10.1016/j.imavis.2019.06.008_bb0035) 2012; 6 Ren (10.1016/j.imavis.2019.06.008_bb0085) 2015 Krizhevsky (10.1016/j.imavis.2019.06.008_bb0045) 2012 LeCun (10.1016/j.imavis.2019.06.008_bb0040) 1998; 86 Tian (10.1016/j.imavis.2019.06.008_bb0015) 2017; 7 Xu (10.1016/j.imavis.2019.06.008_bb0120) 1999; 6 AI Yong-hao (10.1016/j.imavis.2019.06.008_bb0115) 2013; 20 Xu (10.1016/j.imavis.2019.06.008_bb0005) 2002; 3 Paulraj (10.1016/j.imavis.2019.06.008_bb0020) 2010 10.1016/j.imavis.2019.06.008_bb0055 |
| References_xml | – year: 2017 ident: bb0070 article-title: Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning – volume: 86 start-page: 2278 year: 1998 end-page: 2324 ident: bb0040 article-title: Gradient-based learning applied to document recognition [J] publication-title: Proc. IEEE – year: 2016 ident: bb0065 article-title: Deep residual learning for image recognition publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – volume: 20 start-page: 80 year: 2013 end-page: 86 ident: bb0115 article-title: Surface detection of continuous casting slabs based on curvelet transform and kernel locality preserving projections publication-title: J. Iron Steel Res. Int. – volume: 35 start-page: 23 year: 2015 end-page: 30 ident: bb0010 article-title: Application of Shearlet transform to classification of surface defects for metals publication-title: Image Vis. Comput. – volume: 6 start-page: 161 year: 2012 end-page: 166 ident: bb0035 article-title: Steel surface defects detection and classification using SIFT and voting strategy publication-title: Inter. J. of Soft. Eng. and Its Applic. – reference: Lin M., Chen Q., Yan S., 2013. Network in network. arXiv preprinted – reference: Simonyan K., Zisserman A., 2014. Very deep convolutional networks for large-scale image recognition. arXiv preprint – volume: 3 start-page: 028 year: 2002 ident: bb0005 article-title: On-line surface defect inspection system for cold rolled strips [J] publication-title: J. Univ. Sci. Technol. Beijing – volume: 51 start-page: 1 year: 2012 end-page: 8 ident: bb0110 article-title: Feature extraction based on contourlet transform and its application to surface inspection of metals [J] publication-title: Opt. Eng. – year: 2014 ident: bb0100 article-title: Large-scale video classification with convolutional neural networks publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – volume: 7 start-page: 311 year: 2017 ident: bb0015 article-title: An algorithm for surface defect identification of steel plates based on genetic algorithm and extreme learning machine publication-title: Metals – start-page: 585 year: 2012 end-page: 622 ident: bb0030 article-title: Self-organizing maps publication-title: Handbook of Natural Computing – volume: 6 start-page: 296 year: 1999 end-page: 298 ident: bb0120 article-title: Surface inspection system for cold rolled strips based on image processing technique publication-title: Int. J. Miner. Metall. Mater. – year: 2016 ident: bb0090 article-title: You only look once: unified, real-time object detection publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – reference: Redmon, Joseph, Ali Farhadi, 2016. YOLO9000: better, faster, stronger. arXiv preprint – year: 2015 ident: bb0080 article-title: Fast r-cnn publication-title: Proceedings of the IEEE International Conference on Computer Vision – year: 2014 ident: bb0075 article-title: Rich feature hierarchies for accurate object detection and semantic segmentation publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – reference: . – start-page: 1097 year: 2012 end-page: 1105 ident: bb0045 article-title: Imagenet classification with deep convolutional neural networks publication-title: Advances in Neural Information Processing Systems – year: 2013 ident: bb0105 article-title: Sparse autoencoder-based feature transfer learning for speech emotion recognition publication-title: Affective Computing and Intelligent Interaction (ACII), 2013 Humaine Association Conference on IEEE – year: 2010 ident: bb0020 article-title: Structural steel plate damage detection using DFT spectral energy and artificial neural network[C] publication-title: 6th International Colloquium on Signal Processing & its Applications. Malacca – start-page: 91 year: 2015 end-page: 99 ident: bb0085 article-title: Faster R-CNN: towards real-time object detection with region proposal networks publication-title: Advances in Neural Information Processing Systems – volume: 42 start-page: 389 year: 2009 end-page: 397 ident: bb0025 article-title: Automatic detection of cracks in raw steel block using Gabor filter optimized by univariate dynamic encoding algorithm for searches (uDEAS)[J] publication-title: NDT&E International – year: 2015 ident: bb0060 article-title: Going deeper with convolutions publication-title: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition – volume: 51 start-page: 1 issue: 11 year: 2012 ident: 10.1016/j.imavis.2019.06.008_bb0110 article-title: Feature extraction based on contourlet transform and its application to surface inspection of metals [J] publication-title: Opt. Eng. doi: 10.1117/1.OE.51.11.113605 – volume: 35 start-page: 23 year: 2015 ident: 10.1016/j.imavis.2019.06.008_bb0010 article-title: Application of Shearlet transform to classification of surface defects for metals publication-title: Image Vis. Comput. doi: 10.1016/j.imavis.2015.01.001 – volume: 6 start-page: 296 issue: 4 year: 1999 ident: 10.1016/j.imavis.2019.06.008_bb0120 article-title: Surface inspection system for cold rolled strips based on image processing technique publication-title: Int. J. Miner. Metall. Mater. – volume: 7 start-page: 311 issue: 8 year: 2017 ident: 10.1016/j.imavis.2019.06.008_bb0015 article-title: An algorithm for surface defect identification of steel plates based on genetic algorithm and extreme learning machine publication-title: Metals doi: 10.3390/met7080311 – ident: 10.1016/j.imavis.2019.06.008_bb0095 doi: 10.1109/CVPR.2017.690 – year: 2013 ident: 10.1016/j.imavis.2019.06.008_bb0105 article-title: Sparse autoencoder-based feature transfer learning for speech emotion recognition – volume: 3 start-page: 028 year: 2002 ident: 10.1016/j.imavis.2019.06.008_bb0005 article-title: On-line surface defect inspection system for cold rolled strips [J] publication-title: J. Univ. Sci. Technol. Beijing – year: 2014 ident: 10.1016/j.imavis.2019.06.008_bb0075 article-title: Rich feature hierarchies for accurate object detection and semantic segmentation – volume: 42 start-page: 389 issue: 5 year: 2009 ident: 10.1016/j.imavis.2019.06.008_bb0025 article-title: Automatic detection of cracks in raw steel block using Gabor filter optimized by univariate dynamic encoding algorithm for searches (uDEAS)[J] publication-title: NDT&E International doi: 10.1016/j.ndteint.2009.01.007 – year: 2016 ident: 10.1016/j.imavis.2019.06.008_bb0090 article-title: You only look once: unified, real-time object detection – ident: 10.1016/j.imavis.2019.06.008_bb0055 – year: 2016 ident: 10.1016/j.imavis.2019.06.008_bb0065 article-title: Deep residual learning for image recognition – year: 2010 ident: 10.1016/j.imavis.2019.06.008_bb0020 article-title: Structural steel plate damage detection using DFT spectral energy and artificial neural network[C] – start-page: 91 year: 2015 ident: 10.1016/j.imavis.2019.06.008_bb0085 article-title: Faster R-CNN: towards real-time object detection with region proposal networks – volume: 6 start-page: 161 issue: 2 year: 2012 ident: 10.1016/j.imavis.2019.06.008_bb0035 article-title: Steel surface defects detection and classification using SIFT and voting strategy publication-title: Inter. J. of Soft. Eng. and Its Applic. – year: 2017 ident: 10.1016/j.imavis.2019.06.008_bb0070 – year: 2014 ident: 10.1016/j.imavis.2019.06.008_bb0100 article-title: Large-scale video classification with convolutional neural networks – start-page: 1097 year: 2012 ident: 10.1016/j.imavis.2019.06.008_bb0045 article-title: Imagenet classification with deep convolutional neural networks – ident: 10.1016/j.imavis.2019.06.008_bb0050 – start-page: 585 year: 2012 ident: 10.1016/j.imavis.2019.06.008_bb0030 article-title: Self-organizing maps – volume: 86 start-page: 2278 issue: 11 year: 1998 ident: 10.1016/j.imavis.2019.06.008_bb0040 article-title: Gradient-based learning applied to document recognition [J] publication-title: Proc. IEEE doi: 10.1109/5.726791 – volume: 20 start-page: 80 issue: 5 year: 2013 ident: 10.1016/j.imavis.2019.06.008_bb0115 article-title: Surface detection of continuous casting slabs based on curvelet transform and kernel locality preserving projections publication-title: J. Iron Steel Res. Int. doi: 10.1016/S1006-706X(13)60102-8 – year: 2015 ident: 10.1016/j.imavis.2019.06.008_bb0060 article-title: Going deeper with convolutions – year: 2015 ident: 10.1016/j.imavis.2019.06.008_bb0080 article-title: Fast r-cnn |
| SSID | ssj0007079 |
| Score | 2.446755 |
| Snippet | Due to the large intra-class variations and unbalanced training samples, the accuracy of existing algorithms used in defect classification of hot rolled steels... |
| SourceID | crossref elsevier |
| SourceType | Enrichment Source Index Database Publisher |
| StartPage | 12 |
| SubjectTerms | AutoEncoder Convolutional neural networks Defect identification Hot rolled steels Surface inspection |
| Title | Design of multi-scale receptive field convolutional neural network for surface inspection of hot rolled steels |
| URI | https://dx.doi.org/10.1016/j.imavis.2019.06.008 |
| Volume | 89 |
| WOSCitedRecordID | wos000487574900002&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVESC databaseName: ScienceDirect Freedom Collection - Elsevier customDbUrl: eissn: 1872-8138 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0007079 issn: 0262-8856 databaseCode: AIEXJ dateStart: 19950201 isFulltext: true titleUrlDefault: https://www.sciencedirect.com providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Li9swEBYh20N76GPb0u0LHXp1cSzZko9Ld0u3lKXQLeRmZFmCLFlnSZxl--87o5HtlJS-oIc4jogUkfkyGk2--cTYm0zX0uS1SGye2URq2LOWDgK5mdCmMGkNW5agrv9JnZ_r-bz8PJl862thbpaqbfXtbXn9X00NbWBsLJ39C3MPg0ID3IPR4Qpmh-sfGf4kcDLC_-ZIFkw2YAU8GyXwV1DjGzlrgW0epwFGQlXL8BQ44YF6uNmuvbEoKULFmDGuXHVISFxCmArwICnmIbg9u0ICEGbiqWI98NW3Xb84hoxr8HGL_vV8S7VBY1qfXM-JaVaxV8xIzEbKVUyT7ZXKBG-WFeB6dR51r8nbagVtM5J36d0xnSgU_WmkWNPKHKrm9n0-pR8u3y6uUJcB2XplkGRN9bjGDczDLzgPnAYErhBeojzBQabyUk_ZwfHZ6fzjsIyjdCAl6Gjefd1lIAfuf9bP45qdWOXiIbsfNxn8mMDxiE1ce8gexA0Hj-58c8ju7ahRPmYtIYevPN9BDh-QwwNy-A_I4YQcHpHDATk8IoePyMEhATmckMMJOU_Y1_enF-8-JPE4jsTKme6SUmpZ-hoe1jjpskYVPvW5lcrXvnQ-g823MVpKIWpbFE4rJXzaGKFEU6ZOiqds2q5a94zxcIgAbGZVIxvpYHRRFzZrchR39HUjjpjov8zKRq16PDJlWfWkxMuKTFChCarAzdRHLBl6XZNWy2_er3o7VTHepDiyAmj9sufzf-75gt0dfzIv2bRbb90rdsfedIvN-nXE4HeDN6oZ |
| linkProvider | Elsevier |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Design+of+multi-scale+receptive+field+convolutional+neural+network+for+surface+inspection+of+hot+rolled+steels&rft.jtitle=Image+and+vision+computing&rft.au=He%2C+Di&rft.au=Xu%2C+Ke&rft.au=Wang%2C+Dadong&rft.date=2019-09-01&rft.pub=Elsevier+B.V&rft.issn=0262-8856&rft.eissn=1872-8138&rft.volume=89&rft.spage=12&rft.epage=20&rft_id=info:doi/10.1016%2Fj.imavis.2019.06.008&rft.externalDocID=S0262885619300915 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0262-8856&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0262-8856&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0262-8856&client=summon |