[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84063-en":3,"doc-seo-84063-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84063,1374391975076,"Riley","https://ap-avatar.wpscdn.com/avatar/14000253ca4ec9f6853?x-image-process=image/resize,m_fixed,w_180,h_180&k=1783305029341752051",8,"Research & Report","EcoVision AI-Powered Drone Imaging for Salt Marsh Vegetation Monitoring and Dominance Mapping","High-resolution RGB imagery from low-altitude UAV surveys is processed through a modular AI pipeline combining transformer-based semantic segmentation, connected-component vegetation extraction, ConvNeXt fine-grained species classification, and 2×2 m grid-based dominance scoring. The system targets Spartina maritima and Puccinellia maritima, training on curated, manually annotated UAV imagery and biodiversity data from public datasets. Species masks achieve mean IoU ≈0.56 and pixel accuracy ≈0.96, while object-level classification reaches F1 ≈0.99. Dominance estimates align with quadrat field surveys within \u003C8% mean absolute difference, enabling scalable salt marsh monitoring.","arXiv :2607 .06 105v 1 [ cs .CV] 7 Jul 2026  \nEcoVision: AI-Powered Drone Imaging for Salt Marsh Vegetation Monitoring and Dominance Mapping  \nInnocent Onyenonachia , Peter J. Laweranceb , Nadia Kanwala  \na School of Computer Science and Mathematics, Keele University, ST5 3BG, United  \nKingdom  \nb School of Life Science, Keele University, ST5 3BG, United Kingdom  \nAbstract  \nHigh-resolution RGB imagery acquired from low-altitude UAV surveys was processed through a modular pipeline incorporating transformer-based semantic segmentation, connected-component vegetation extraction, fine-grained species classification using a ConvNeXt architecture, and grid-based dominance scoring at 2×2m resolution. The framework targeted two ecologically significant halophytic grasses, Spartina maritima and Puccinellia maritima, and was trained using a curated and manually annotated UAV imagery, along with biodiversity imagery sourced from publicly accessible datasets. In order to identify these plants from the imagery, our segmentation yielded reliable species masks (mean IoU ≈ 0.56; pixel-level accuracy ≈ 0.96), while objectlevel classification achieved very good discrimination (F1 ≈ 0.99) . Dominance estimates closely matched quadrat-based field surveys, with mean absolute differences below 8%, preserving fine-scale spatial structure under realistic survey conditions. The developed system, named EcoVision, establishes a practical foundation for scalable, high-resolution salt marsh monitoring, demonstrating how AI-driven workflows can translate pixel-level predictions into ecologically interpretable metrics.  \nKeywords: method, dataset, pipeline, segform, convnext, training, preprocessing, annotation, augment, collection, uav survey, data split  \n1. Introduction  \nAmong coastal landscapes, vegetation uniquely underpins ecosystem service delivery and functioning, regulating biodiversity patterns, habitat structure, and key ecological processes [1] . Accurate, high-resolution monitoring is  \ncritical for assessing environmental change, guiding conservation strategies, and informing predictive ecological models [2, 3] . Despite their long-standing role in ecology, field surveys depend heavily on human judgement and manual effort, introducing variability and limiting scalability for both temporal and spatial coverage [4, 5] .  \nThese constraints are particularly pronounced in dynamic or inaccessible habitats, such as salt marshes, where interestingly species of quite different traits and morphology can co-exist in physical niches defined by mere tens of centimetres. The consequences of these very strong physical niches, yet also expansive, inaccessible, and heterogeneous settings, mean traditional surveys are both onerous, focused typically on transects, and often generalise a marsh rather than perhaps finding specific areas of concern that might be missed by linear walks [6] . Satellite imagery might counter this limitation; however, satellite data lacks the spatial resolution to identify species and quantify their associated ecosystem service value [7], and it also lacks the temporal flexibility necessary for species-level analysis [8, 9] . Emerging UAV-based remote sensing offers high-resolution, repeatable imaging capabilities that capture fine-scale vegetation patterns. At the same time, machine learning methods enable the automated identification, segmentation, and dominance assessment of species [10] . By combining UAV imagery with computer vision and deep learning, the proposed framework addresses critical limitations of conventional monitoring, providing scalable, objective, and ecologically interpretable data. This integration also supports further advances such as near real-time ecosystem assessment, invasive species management, and evidencebased conservation planning, bridging the gap between high-resolution imagery and actionable ecological insight [11, 12, 6] .  \n1.1. Research Objectives  \nThis research aims to develop and evaluate an ","cbCaid1ZhEbB7SeH","https://ap.wps.com/l/cbCaid1ZhEbB7SeH","pdf",5075206,1,37,"English","en",105,"# Introduction\n## Research Objectives\n## Contributions of This Study","[{\"question\":\"What is EcoVision designed to measure in salt marshes?\",\"answer\":\"EcoVision targets automated plant species identification and dominance assessment for two halophytic grasses in salt marsh ecosystems, producing ecologically interpretable dominance scores at a 2×2 m patch scale.\"},{\"question\":\"How does the pipeline turn UAV images into dominance metrics?\",\"answer\":\"It uses transformer-based semantic segmentation for vegetation and species masks, blob-level/object-level classification with a ConvNeXt architecture, and then aggregates predictions into quantitative dominance scores on a 2×2 m grid.\"},{\"question\":\"How accurate are the model outputs compared with field surveys?\",\"answer\":\"Segmentation yields mean IoU around 0.56 with pixel-level accuracy about 0.96, classification reaches F1 about 0.99, and dominance estimates closely match quadrat-based surveys with mean absolute differences below 8%.\"}]",1784192329,93,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"ecovision-ai-powered-drone-imaging-for-salt-marsh-vegetation-monitoring-and-dominance-mapping","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/ecovision-ai-powered-drone-imaging-for-salt-marsh-vegetation-monitoring-and-dominance-mapping/84063/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is EcoVision designed to measure in salt marshes?","Question",{"text":75,"@type":76},"EcoVision targets automated plant species identification and dominance assessment for two halophytic grasses in salt marsh ecosystems, producing ecologically interpretable dominance scores at a 2×2 m patch scale.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the pipeline turn UAV images into dominance metrics?",{"text":80,"@type":76},"It uses transformer-based semantic segmentation for vegetation and species masks, blob-level/object-level classification with a ConvNeXt architecture, and then aggregates predictions into quantitative dominance scores on a 2×2 m grid.",{"name":82,"@type":73,"acceptedAnswer":83},"How accurate are the model outputs compared with field surveys?",{"text":84,"@type":76},"Segmentation yields mean IoU around 0.56 with pixel-level accuracy about 0.96, classification reaches F1 about 0.99, and dominance estimates closely match quadrat-based surveys with mean absolute differences below 8%.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]