YOLOv8-GK: Tomato leaf disease detection algorithm

YOLOv8-GK: Tomato leaf disease detection algorithm | 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 Article YOLOv8-GK: Tomato leaf disease detection algorithm Huiting Xiao, Chongyang Ning, Haojie He This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5454516/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 Tomato leaf diseases seriously affect the yield and quality of tomatoes, so it is crucial to detect leaf diseases accurately. In recent years, the use of deep learning techniques has significantly improved the accuracy of plant leaf disease detection. However, the existing deep learning detection models perform poorly when confronted with the mutual occlusion between leaves, the change of light, and interference from complex backgrounds such as plants and soil. To overcome the challenge of accurately locating and classifying tomato leaf diseases in complex scenarios, this paper proposes a real-time leaf disease detection method based on YOLOv8. On the one hand, to address the defect of capturing global contextual information and the problem of gradient vanishing in deep networks, a GIP-SPPF module is constructed by integrating global background information and edge information. This module preserves the edge cross information, effectively reducing the training difficulty while improving the model performance. On the other hand, to effectively reduce the size of model parameters, we introduced a depthwise separable convolutional structure and combined it with the idea of a residual connection mechanism to construct KOIDSConv, which not only significantly improves the lightweight of the model, but also enhances feature fusion and information complementarity between different channels. The results show that the YOLOv8-GK model has better detection performance than other detection methods in the dataset constructed in this paper, with a high mAP of 98.06%. Deep learning Plant disease detection Depthwise separable convolution SPPF Full Text Additional Declarations No competing interests reported. Supplementary Files Fig.S1.png Fig.S2d1.png Fig.S3.png Fig.S4.png Fig.S5.png Fig.S6.jpg Fig.S7.png Fig.S8.png Fig.S9.png 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-5454516","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":379045459,"identity":"1031dc64-1031-4165-ab9e-ec123b9bbc10","order_by":0,"name":"Huiting Xiao","email":"","orcid":"","institution":"Central South University of Forestry and Technology","correspondingAuthor":false,"prefix":"","firstName":"Huiting","middleName":"","lastName":"Xiao","suffix":""},{"id":379045460,"identity":"d7d2c456-afcf-44da-908b-16147a0031fd","order_by":1,"name":"Chongyang Ning","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYJCCAwwMNkCK8QGQYCZaSxpItWED0VqA4DAJWvhn5B48XPDrvLz8jGT2BwwV1okN7GcP4NUicSMv4fDMvtuGG24kMzYwnElPbODJS8CrxUAix+Awb8/tBAOJ/IMNjG2HExskeAyI0XIuAegwxgbGf8Rq4flxIIEB5DDGBiK0SJx5A7SlIdlww5nHjDMSjqUbt/Hk4NfC355j/Jnnj528fHsyw4cPNday/exn8GthEEgAxnsblANkM7DhVw+y5gCQ+ENQ2SgYBaNgFIxkAABOwkcfJ0oCAwAAAABJRU5ErkJggg==","orcid":"","institution":"Central South University of Forestry and Technology","correspondingAuthor":true,"prefix":"","firstName":"Chongyang","middleName":"","lastName":"Ning","suffix":""},{"id":379045461,"identity":"9da3cc82-9d30-4092-b8cd-48cd5b0628c5","order_by":2,"name":"Haojie He","email":"","orcid":"","institution":"Central South University of Forestry and Technology","correspondingAuthor":false,"prefix":"","firstName":"Haojie","middleName":"","lastName":"He","suffix":""}],"badges":[],"createdAt":"2024-11-14 14:08:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5454516/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5454516/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83713319,"identity":"5b02f597-0a25-427f-972e-8eaede4f4fe6","added_by":"auto","created_at":"2025-05-31 17:23:49","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2619209,"visible":true,"origin":"","legend":"","description":"","filename":"LaTeX.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1_covered_efeaab50-57c9-41d3-80eb-75e6933cf18e.pdf"},{"id":71317004,"identity":"1f49e193-9c52-431a-ae3e-7aa1ed209736","added_by":"auto","created_at":"2024-12-13 09:18:24","extension":"png","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":530116,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S1.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/0e048e0163f5de636eca5b3f.png"},{"id":71317007,"identity":"7c1ddc8b-9473-4621-9fd0-171882a2e099","added_by":"auto","created_at":"2024-12-13 09:18:24","extension":"png","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":2873008,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S2d1.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/8ebd2ceb48f0b335d5e683be.png"},{"id":71317010,"identity":"47bfd9d4-3968-4002-b790-4bbfb421714d","added_by":"auto","created_at":"2024-12-13 09:18:24","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":568432,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S3.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/693eb49b2bb0c067e18bb183.png"},{"id":71317619,"identity":"a3dbcf03-4650-4968-97e6-999eb593d3d4","added_by":"auto","created_at":"2024-12-13 09:26:24","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":44978,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S4.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/7e5f26a87fd725bca14542b7.png"},{"id":71318087,"identity":"c07d2480-366b-4b68-984b-6f835b6b1c4d","added_by":"auto","created_at":"2024-12-13 09:34:24","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":118761,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S5.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/10e07ef1d3242a7a3866f8c6.png"},{"id":71317618,"identity":"b2de3c90-7c86-4d52-a5ed-483d024e30dc","added_by":"auto","created_at":"2024-12-13 09:26:24","extension":"jpg","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":37886,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/222b814116dce992fe54f131.jpg"},{"id":71317620,"identity":"28ed875f-2179-4656-980c-33058f03c10e","added_by":"auto","created_at":"2024-12-13 09:26:24","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":51384,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S7.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/961e45e01ce679fba3fa44e0.png"},{"id":71317012,"identity":"9f522d02-df80-4078-a9c6-658ef6b518bd","added_by":"auto","created_at":"2024-12-13 09:18:24","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":2916685,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S8.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/8f607c2609ec9108da0a2804.png"},{"id":71317011,"identity":"6129957a-167f-4674-8cae-c019d3f2bfa1","added_by":"auto","created_at":"2024-12-13 09:18:24","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":4530311,"visible":true,"origin":"","legend":"","description":"","filename":"Fig.S9.png","url":"https://assets-eu.researchsquare.com/files/rs-5454516/v1/2cbb570d3c57e0214154c0f7.png"}],"financialInterests":"No competing interests reported.","formattedTitle":"YOLOv8-GK: Tomato leaf disease detection algorithm","fulltext":[],"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":true,"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":"Deep learning, Plant disease detection, Depthwise separable convolution, SPPF","lastPublishedDoi":"10.21203/rs.3.rs-5454516/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5454516/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Tomato leaf diseases seriously affect the yield and quality of tomatoes, so it is crucial to detect leaf diseases accurately. In recent years, the use of deep learning techniques has significantly improved the accuracy of plant leaf disease detection. However, the existing deep learning detection models perform poorly when confronted with the mutual occlusion between leaves, the change of light, and interference from complex backgrounds such as plants and soil. To overcome the challenge of accurately locating and classifying tomato leaf diseases in complex scenarios, this paper proposes a real-time leaf disease detection method based on YOLOv8. On the one hand, to address the defect of capturing global contextual information and the problem of gradient vanishing in deep networks, a GIP-SPPF module is constructed by integrating global background information and edge information. This module preserves the edge cross information, effectively reducing the training difficulty while improving the model performance. On the other hand, to effectively reduce the size of model parameters, we introduced a depthwise separable convolutional structure and combined it with the idea of a residual connection mechanism to construct KOIDSConv, which not only significantly improves the lightweight of the model, but also enhances feature fusion and information complementarity between different channels. The results show that the YOLOv8-GK model has better detection performance than other detection methods in the dataset constructed in this paper, with a high mAP of 98.06%. ","manuscriptTitle":"YOLOv8-GK: Tomato leaf disease detection algorithm","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-13 09:18:19","doi":"10.21203/rs.3.rs-5454516/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":"a1f49f79-bea4-4a2e-a063-7343502873e4","owner":[],"postedDate":"December 13th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-05-31T17:23:28+00:00","versionOfRecord":[],"versionCreatedAt":"2024-12-13 09:18:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5454516","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5454516","identity":"rs-5454516","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}