Rapid and Accurate Varieties Identification of Different Crop Seeds Under Sample-Limited Condition Based on Hyperspectral Imaging and Deep Transfer Learning

Rapid and Accurate Varieties Identification of Different Crop Seeds Under Sample-Limited Condition Based on Hyperspectral Imaging and Deep Transfer Learning | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Rapid and Accurate Varieties Identification of Different Crop Seeds Under Sample-Limited Condition Based on Hyperspectral Imaging and Deep Transfer Learning Na Wu, Fei Liu, Yidan Bao, Mu Li, Wei Huang, Fanjia Meng, Chu Zhang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-108818/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Varieties identification of crop seeds is significant for breeders to screen out seeds with specific traits and for market regulators to detect seeds purity. Hyperspectral imaging technology provides a fast and non-destructive means for varieties identification. And deep learning algorithm is suitable for effective analysis of redundant spectral data. However, deep learning algorithms have serious big data dependency, while collecting high-quality large-scale samples was high-cost in many cases. This made it difficult to build an accurate identification model. This study aimed to explore a rapid and accurate method for varieties identification of different crop seeds under sample-limited condition based on hyperspectral imaging and deep transfer learning. Results: Three deep neural networks with typical structures were designed based on a samples-rich Pea dataset. Obtained the highest accuracy of 99.57 %, VGG-MODEL was transferred to classify four target datasets (Rice, Oat, Wheat, Cotton) with limited samples. The accuracies of deep transferred model achieved 95 %, 99 %, 80.8 %, and 83.86 % on the four datasets, respectively. Using training sets with different sizes, deep transferred model could always obtain higher performance than other traditional methods. Visualization of training process and classification results confirmed the portability of common features of seed spectra and provided an interpreted method for rapid and accurate varieties identification of crop seeds. Conclusions: This study combined hyperspectral imaging and deep transfer learning to identify varieties of different crop seeds, which was proved to be efficient under sample-limited condition. This facilitated crop variety screening process under the scenario of sample scarcity. It also provided a new idea for the detection of other qualities of crop seeds based on hyperspectral imaging under sample-limited condition. Plant Molecular Biology and Genetics Plant Physiology and Morphology Hyperspectral imaging Crop seeds Deep learning Transfer learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Full Text Supplementary Files Additionalfile1TableS1.docx Additional file 1: Table S1. The detailed parameters of three deep neural networks. Additionalfile2Fig.S1.tif Additional file 2: Fig. S1. Feature visualization of VGG-MODEL on Rice dataset using t-SNE. Additionalfile3Fig.S2.tif Additional file 3: Fig. S2. Feature visualization of VGG-MODEL on Oat dataset using t-SNE. Additionalfile4Fig.S3.tif Additional file 4: Fig. S3. Feature visualization of VGG-MODEL on Wheat dataset using t-SNE. Additionalfile5Fig.S4.tif Additional file 5: Fig. S4. Feature visualization of VGG-MODEL on Cotton dataset using t-SNE. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-108818","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research","associatedPublications":[],"authors":[{"id":4795397,"identity":"fc513153-75f8-4ffc-85bd-a161b37c4226","order_by":0,"name":"Na Wu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Na","middleName":"","lastName":"Wu","suffix":""},{"id":4795398,"identity":"e0cfa4af-3302-483f-a7df-ae6d83954000","order_by":1,"name":"Fei Liu","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Fei","middleName":"","lastName":"Liu","suffix":""},{"id":4795399,"identity":"2dc0771a-4a1a-435d-8361-a6c1f9584803","order_by":2,"name":"Yidan Bao","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Yidan","middleName":"","lastName":"Bao","suffix":""},{"id":4795400,"identity":"b8c499f1-c4a7-4e7e-b150-f719a7f9b6db","order_by":3,"name":"Mu Li","email":"","orcid":"","institution":"Jilin Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Mu","middleName":"","lastName":"Li","suffix":""},{"id":4795401,"identity":"5575e9e7-1abc-4e87-8621-fca9e96ad2ff","order_by":4,"name":"Wei Huang","email":"","orcid":"","institution":"Jilin Academy of Agricultural Sciences","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Huang","suffix":""},{"id":4795402,"identity":"472553a9-c32f-4b19-bda8-9a35199618ac","order_by":5,"name":"Fanjia Meng","email":"","orcid":"","institution":"China Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Fanjia","middleName":"","lastName":"Meng","suffix":""},{"id":4795403,"identity":"0eb55088-d861-49ac-9e2f-8cf3aafe91e9","order_by":6,"name":"Chu Zhang","email":"","orcid":"","institution":"Zhejiang University","correspondingAuthor":false,"prefix":"","firstName":"Chu","middleName":"","lastName":"Zhang","suffix":""},{"id":4795404,"identity":"59e8e73a-3099-40ba-971e-637b92238518","order_by":7,"name":"Yong He","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAArklEQVRIiWNgGAWjYDACCQbGBww8IFYC8VqYDUjWwiYBYRGrxeB287HKHzKHGfjZcwwYfu4gRsudY2k3JHgOM0j2vDFg7D1DjJYbOWY3DIBagAwDZsY2orTkfytIAGqxJ0FLDhvDAZAtEsRqkbyRZizZwJPOI3HmWcHBXmK08N1IfvjxZ4+1HH978sYHP4nRonAASDD2QCLzABEaGBjkG0DkD6LUjoJRMApGwUgFAKARNL2SG7lGAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-6752-1757","institution":"Zhejiang University","correspondingAuthor":true,"prefix":"","firstName":"Yong","middleName":"","lastName":"He","suffix":""}],"badges":[],"createdAt":"2020-11-15 23:42:49","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-108818/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-108818/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":3724389,"identity":"7ba54540-5f4e-40f5-bd8d-616b6809d49a","added_by":"auto","created_at":"2020-11-20 15:43:35","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":31652,"visible":true,"origin":"","legend":"The structures of three developed deep neural networks","description":"","filename":"OnlineFig.1.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/a096b6b6da5ea79e52e244d2.Png"},{"id":3724391,"identity":"0a8c12e3-e2ab-4a28-9cd3-5dc68fc7539a","added_by":"auto","created_at":"2020-11-20 15:43:35","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":23276,"visible":true,"origin":"","legend":"The inner structures of three typical blocks","description":"","filename":"OnlineFig.2.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/e908825ae88f4cb63098f809.Png"},{"id":3724393,"identity":"67e77c17-b394-455b-ae47-679f8b4dcceb","added_by":"auto","created_at":"2020-11-20 15:43:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":24711,"visible":true,"origin":"","legend":"Transfer learning strategy. a The common process of transfer learning. b The deep transfer learning strate-gy in this study","description":"","filename":"OnlineFig.3.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/44ed83ec15b98e7556451a24.Png"},{"id":3724395,"identity":"c3284a99-cf93-4d63-93c8-c57b25b5677b","added_by":"auto","created_at":"2020-11-20 15:43:36","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":449537,"visible":true,"origin":"","legend":"The average spectra with standard deviation of five crop seeds","description":"","filename":"OnlineFig.4.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/4eff9f17d48a24e38e581e5b.Png"},{"id":3724397,"identity":"1eaab348-f410-40ec-92fc-d61a6f00e956","added_by":"auto","created_at":"2020-11-20 15:43:37","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1780381,"visible":true,"origin":"","legend":"Classification accuracies of all models on the four target datasets","description":"","filename":"OnlineFig.5.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/86261731fb2a62e2e6857faa.Png"},{"id":3724398,"identity":"2283a93f-7807-4620-8e7f-a7966fe1b3fb","added_by":"auto","created_at":"2020-11-20 15:43:37","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":254683,"visible":true,"origin":"","legend":"The accuracies and losses of VGG-MODEL on the training and validation set","description":"","filename":"OnlineFig.6.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/609146990015ceaa51d18c60.Png"},{"id":3724399,"identity":"801b86ca-332e-4001-967c-9c2bac9347b5","added_by":"auto","created_at":"2020-11-20 15:43:38","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":612218,"visible":true,"origin":"","legend":"Feature visualization of VGG-MODEL on Pea dataset using t-SNE","description":"","filename":"OnlineFig.7.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/2637e85cf4eb82290eda36d3.Png"},{"id":3724400,"identity":"0d3fd436-3aa2-4f57-bb69-1a8ab35d7839","added_by":"auto","created_at":"2020-11-20 15:43:38","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":119500,"visible":true,"origin":"","legend":"Classification maps of pea seeds","description":"","filename":"OnlineFig.8.Png","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/4fb60050b005b933412fee53.Png"},{"id":13557557,"identity":"d99f8c94-f3cd-4b87-8f1a-307fb488f4e2","added_by":"auto","created_at":"2021-09-17 02:53:03","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1270055,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscritptPM.pdf","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1_covered.pdf"},{"id":3724387,"identity":"0af14812-4b0c-4c77-9365-1a6b3cf7f3ad","added_by":"auto","created_at":"2020-11-20 15:43:38","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":737339,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscritptPM.pdf","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1_stamped.pdf"},{"id":3724388,"identity":"a16e8a3e-1e5a-473b-b8c6-5abf8daf29a8","added_by":"auto","created_at":"2020-11-20 15:43:35","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":17867,"visible":true,"origin":"","legend":"Additional file 1: Table S1. The detailed parameters of three deep neural networks.","description":"","filename":"Additionalfile1TableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/fe8937ebb6b10f6d5437eda3.docx"},{"id":3724390,"identity":"247808d5-97a1-40c9-b053-acfb3680baab","added_by":"auto","created_at":"2020-11-20 15:43:35","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":236080,"visible":true,"origin":"","legend":"Additional file 2: Fig. S1. Feature visualization of VGG-MODEL on Rice dataset using t-SNE.","description":"","filename":"Additionalfile2Fig.S1.tif","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/5a2be6957f066129c711b75e.tif"},{"id":3724392,"identity":"8098fc47-0871-408f-b7b3-8d44e4827336","added_by":"auto","created_at":"2020-11-20 15:43:36","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":264684,"visible":true,"origin":"","legend":"Additional file 3: Fig. S2. Feature visualization of VGG-MODEL on Oat dataset using t-SNE.","description":"","filename":"Additionalfile3Fig.S2.tif","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/8032d74f6b303fb69ab2abc6.tif"},{"id":3724394,"identity":"fdbf9d2a-1062-4064-95ce-093b999d48d3","added_by":"auto","created_at":"2020-11-20 15:43:36","extension":"tif","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":339652,"visible":true,"origin":"","legend":"Additional file 4: Fig. S3. Feature visualization of VGG-MODEL on Wheat dataset using t-SNE.","description":"","filename":"Additionalfile4Fig.S3.tif","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/c8891cb5dcbadd749b461f37.tif"},{"id":3724396,"identity":"05bab1f5-fa2f-4dd8-9876-4d8cf823c922","added_by":"auto","created_at":"2020-11-20 15:43:37","extension":"tif","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":408016,"visible":true,"origin":"","legend":"Additional file 5: Fig. S4. Feature visualization of VGG-MODEL on Cotton dataset using t-SNE.","description":"","filename":"Additionalfile5Fig.S4.tif","url":"https://assets-eu.researchsquare.com/files/rs-108818/v1/c590f1dce32a5e905903ce2a.tif"}],"financialInterests":"","formattedTitle":"\u003cp\u003eRapid and Accurate Varieties Identification of Different Crop Seeds Under Sample-Limited Condition Based on Hyperspectral Imaging and Deep Transfer Learning\u003c/p\u003e","fulltext":[{"header":"Full Text","content":"\u003cp\u003eThis preprint is available for \u003ca href='/article/rs-108818/latest.pdf' target='_blank'\u003edownload as a PDF\u003c/a\u003e.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Hyperspectral imaging, Crop seeds, Deep learning, Transfer learning","lastPublishedDoi":"10.21203/rs.3.rs-108818/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-108818/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Varieties identification of crop seeds is significant for breeders to screen out seeds with specific traits and for market regulators to detect seeds purity. Hyperspectral imaging technology provides a fast and non-destructive means for varieties identification. And deep learning algorithm is suitable for effective analysis of redundant spectral data. However, deep learning algorithms have serious big data dependency, while collecting high-quality large-scale samples was high-cost in many cases. This made it difficult to build an accurate identification model. This study aimed to explore a rapid and accurate method for varieties identification of different crop seeds under sample-limited condition based on hyperspectral imaging and deep transfer learning.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eResults:\u003c/strong\u003e Three deep neural networks with typical structures were designed based on a samples-rich Pea dataset. Obtained the highest accuracy of 99.57 %, VGG-MODEL was transferred to classify four target datasets (Rice, Oat, Wheat, Cotton) with limited samples. The accuracies of deep transferred model achieved 95 %, 99 %, 80.8 %, and 83.86 % on the four datasets, respectively. Using training sets with different sizes, deep transferred model could always obtain higher performance than other traditional methods. Visualization of training process and classification results confirmed the portability of common features of seed spectra and provided an interpreted method for rapid and accurate varieties identification of crop seeds.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e This study combined hyperspectral imaging and deep transfer learning to identify varieties of different crop seeds, which was proved to be efficient under sample-limited condition. This facilitated crop variety screening process under the scenario of sample scarcity. It also provided a new idea for the detection of other qualities of crop seeds based on hyperspectral imaging under sample-limited condition.\u003c/p\u003e","manuscriptTitle":"Rapid and Accurate Varieties Identification of Different Crop Seeds Under Sample-Limited Condition Based on Hyperspectral Imaging and Deep Transfer Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2020-11-20 15:43:33","doi":"10.21203/rs.3.rs-108818/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"3cdc7c9e-6f8e-4dd6-867b-568aae5fbf0e","owner":[],"postedDate":"November 20th, 2020","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":1151448,"name":"Plant Molecular Biology and Genetics"},{"id":1151449,"name":"Plant Physiology and Morphology"}],"tags":[],"updatedAt":"2021-01-12T14:27:20+00:00","versionOfRecord":[],"versionCreatedAt":"2020-11-20 15:43:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-108818","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-108818","identity":"rs-108818","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}