{"id":{"repo_id":"vilnius","oai_identifier":"oai:vu.lt:elaba:308338957"},"canonical_url":"https://search.ndltd.org/etd/vilnius/oai:vu.lt:elaba:308338957","repository":{"repo_id":"vilnius","name":"Vilnius University","base_url":"https://epublications.vu.lt/oai"},"display":{"title":"Neuroninių tinklų taikymas mikroplastikų segmentavimui /","abstract":"The problem of microplastic pollution is becoming increasingly relevant to society, but the longterm impact of microplastic particles on human health and ecosystems has still not been fully studied. Accurate assessment of microplastic impact requires effective identification methods, however, traditional optical techniques are slow, expensive, and have limited applicability to large volumes of data. Deep learning methods offer the possibility of automating the microplastic identification process, while image preprocessing methods may influence its results. In this work, the „MNv4-Conv-M-fpn” model was applied to microplastic segmentation using the FTMC-provided dataset and the publicly available MDPI dataset. A study of the influence of nine image preprocessing methods was also conducted across both datasets. The experimental results were compared with one another and with models evaluated in the literature review, and the key factors influencing segmentation quality were discussed. The suitability of image preprocessing methods across different datasets was also analysed. On the FTMC dataset, an F1 score of 71.91% was achieved, with the best pre-processing methods improving the baseline result by up to 2.47%. On the MDPI dataset, an F1 score of 52.53% was achieved, however, pre-processing methods did not improve microplastics segmentation performance.","abstract_html":"The problem of microplastic pollution is becoming increasingly relevant to society, but the longterm impact of microplastic particles on human health and ecosystems has still not been fully studied. Accurate assessment of microplastic impact requires effective identification methods, however, traditional optical techniques are slow, expensive, and have limited applicability to large volumes of data. Deep learning methods offer the possibility of automating the microplastic identification process, while image preprocessing methods may influence its results. In this work, the „MNv4-Conv-M-fpn” model was applied to microplastic segmentation using the FTMC-provided dataset and the publicly available MDPI dataset. A study of the influence of nine image preprocessing methods was also conducted across both datasets. The experimental results were compared with one another and with models evaluated in the literature review, and the key factors influencing segmentation quality were discussed. The suitability of image preprocessing methods across different datasets was also analysed. On the FTMC dataset, an F1 score of 71.91% was achieved, with the best pre-processing methods improving the baseline result by up to 2.47%. On the MDPI dataset, an F1 score of 52.53% was achieved, however, pre-processing methods did not improve microplastics segmentation performance.","abstract_has_math":false,"creators":["Švager, Daniil,"],"institution":"Institutional Repository of Vilnius University","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-19T10:12:59Z","subjects":[],"languages":["lit"],"rights":["info:eu-repo/semantics/openAccess"],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://repository.vu.lt/VU:ELABAETD308338957&prefLang=en_US","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://repository.vu.lt/VU:ELABAETD308338957&prefLang=en_US","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":"open","embargo_until":null,"access_note":null},"source_record":{"url":"https://epublications.vu.lt/oai?verb=GetRecord&metadataPrefix=oai_dc&identifier=oai%3Avu.lt%3Aelaba%3A308338957","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Švager, Daniil,"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2026"]},{"key":"dc:publisher","label":"Institution","values":["Institutional Repository of Vilnius University"]},{"key":"dc:relation","label":"Dc Relation","values":["https://epublications.vu.lt/object/elaba:308338957/308338957.pdf"]},{"key":"dc:type","label":"Dc Type","values":["info:eu-repo/semantics/bachelorThesis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["lit"]},{"key":"dc:rights","label":"Dc Rights","values":["info:eu-repo/semantics/openAccess"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://repository.vu.lt/VU:ELABAETD308338957&prefLang=en_US"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["The problem of microplastic pollution is becoming increasingly relevant to society, but the longterm impact of microplastic particles on human health and ecosystems has still not been fully studied. Accurate assessment of microplastic impact requires effective identification methods, however, traditional optical techniques are slow, expensive, and have limited applicability to large volumes of data. Deep learning methods offer the possibility of automating the microplastic identification process, while image preprocessing methods may influence its results. In this work, the „MNv4-Conv-M-fpn” model was applied to microplastic segmentation using the FTMC-provided dataset and the publicly available MDPI dataset. A study of the influence of nine image preprocessing methods was also conducted across both datasets. The experimental results were compared with one another and with models evaluated in the literature review, and the key factors influencing segmentation quality were discussed. The suitability of image preprocessing methods across different datasets was also analysed. On the FTMC dataset, an F1 score of 71.91% was achieved, with the best pre-processing methods improving the baseline result by up to 2.47%. On the MDPI dataset, an F1 score of 52.53% was achieved, however, pre-processing methods did not improve microplastics segmentation performance."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Neuroninių tinklų taikymas mikroplastikų segmentavimui /","Application of neural networks for microplastics segmentation."]}]}],"canonical_facts":{"dc:creator":["Švager, Daniil,"],"dc:date":["2026"],"dc:description":["The problem of microplastic pollution is becoming increasingly relevant to society, but the longterm impact of microplastic particles on human health and ecosystems has still not been fully studied. Accurate assessment of microplastic impact requires effective identification methods, however, traditional optical techniques are slow, expensive, and have limited applicability to large volumes of data. Deep learning methods offer the possibility of automating the microplastic identification process, while image preprocessing methods may influence its results. In this work, the „MNv4-Conv-M-fpn” model was applied to microplastic segmentation using the FTMC-provided dataset and the publicly available MDPI dataset. A study of the influence of nine image preprocessing methods was also conducted across both datasets. The experimental results were compared with one another and with models evaluated in the literature review, and the key factors influencing segmentation quality were discussed. The suitability of image preprocessing methods across different datasets was also analysed. On the FTMC dataset, an F1 score of 71.91% was achieved, with the best pre-processing methods improving the baseline result by up to 2.47%. On the MDPI dataset, an F1 score of 52.53% was achieved, however, pre-processing methods did not improve microplastics segmentation performance."],"dc:format":["application/pdf"],"dc:identifier":["https://repository.vu.lt/VU:ELABAETD308338957&prefLang=en_US"],"dc:language":["lit"],"dc:publisher":["Institutional Repository of Vilnius University"],"dc:relation":["https://epublications.vu.lt/object/elaba:308338957/308338957.pdf"],"dc:rights":["info:eu-repo/semantics/openAccess"],"dc:title":["Neuroninių tinklų taikymas mikroplastikų segmentavimui /","Application of neural networks for microplastics segmentation."],"dc:type":["info:eu-repo/semantics/bachelorThesis"]},"updated_at":"2026-11-19T10:12:59Z"}