Simple item page
 

Comparative Evaluation of CNN and Transformer Architectures for Flowering Phase Classification of Tilia cordata Mill. with Automated Image Quality Filtering

cris.virtual.author-orcid0000-0002-9012-4060
cris.virtual.author-orcid0000-0002-2222-6866
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid0000-0003-3721-6473
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orciddb2bbd84-5a3b-4eb8-ad6c-5aed8ba9e1c9
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid09c81af2-c2b2-424c-b2ec-4cda600883bf
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid3fe42726-36c4-478a-818f-a10f72d4a6ef
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.author-orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
dc.abstract.enUnderstanding and monitoring the phenological phases of trees is essential for ecological research and climate change studies. In this work, we present a comprehensive evaluation of state-of-the-art convolutional neural networks (CNNs) and transformer architectures for the automated classification of the flowering phase of Tilia cordata Mill. (small-leaved lime) based on a large set of real-world images acquired under natural field conditions. The study introduces a novel, automated image quality filtering approach using an XGBoost classifier trained on diverse exposure and sharpness features to ensure robust input data for subsequent deep learning models. Seven modern neural network architectures, including VGG16, ResNet50, EfficientNetB3, MobileNetV3 Large, ConvNeXt Tiny, Vision Transformer (ViT-B/16), and Swin Transformer Tiny, were fine-tuned and evaluated under a rigorous cross-validation protocol. All models achieved excellent performance, with cross-validated F1-scores exceeding 0.97 and balanced accuracy up to 0.993. The best results were obtained for ResNet50 and ConvNeXt Tiny (F1-score: 0.9879 ± 0.0077 and 0.9860 ± 0.0073, balanced accuracy: 0.9922 ± 0.0054 and 0.9927 ± 0.0042, respectively), indicating outstanding sensitivity and specificity for both flowering and non-flowering classes. Classical CNNs (VGG16, ResNet50, and ConvNeXt Tiny) demonstrated slightly superior robustness compared to transformer-based models, though all architectures maintained high generalization and minimal variance across folds. The integrated quality assessment and classification pipeline enables scalable, high-throughput monitoring of flowering phases in natural environments. The proposed methodology is adaptable to other plant species and locations, supporting future ecological monitoring and climate studies. Our key contributions are as follows: (i) introducing an automated exposure-quality filtering stage for field imagery; (ii) publishing a curated, season-long dataset of Tilia cordata images; and (iii) providing the first systematic cross-validated benchmark that contrasts classical CNNs with transformer architectures for phenological phase recognition.
dc.affiliationWydział Inżynierii Środowiska i Inżynierii Mechanicznej
dc.affiliation.instituteKatedra Ekologii i Ochrony Środowiska
dc.affiliation.instituteKatedra Inżynierii Biosystemów
dc.contributor.authorArct, Bogdan
dc.contributor.authorŚwiderski, Bartosz
dc.contributor.authorRóżańska , Monika A.
dc.contributor.authorChojnicki, Bogdan
dc.contributor.authorWojciechowski, Tomasz
dc.contributor.authorNiedbała, Gniewko
dc.contributor.authorKruk, Michał
dc.contributor.authorBobran, Krzysztof
dc.contributor.authorKurek, Jarosław
dc.date.access2025-08-29
dc.date.accessioned2025-08-29T07:50:45Z
dc.date.available2025-08-29T07:50:45Z
dc.date.copyright2025-08-27
dc.date.issued2025
dc.description.abstract<jats:p>Understanding and monitoring the phenological phases of trees is essential for ecological research and climate change studies. In this work, we present a comprehensive evaluation of state-of-the-art convolutional neural networks (CNNs) and transformer architectures for the automated classification of the flowering phase of Tilia cordata Mill. (small-leaved lime) based on a large set of real-world images acquired under natural field conditions. The study introduces a novel, automated image quality filtering approach using an XGBoost classifier trained on diverse exposure and sharpness features to ensure robust input data for subsequent deep learning models. Seven modern neural network architectures, including VGG16, ResNet50, EfficientNetB3, MobileNetV3 Large, ConvNeXt Tiny, Vision Transformer (ViT-B/16), and Swin Transformer Tiny, were fine-tuned and evaluated under a rigorous cross-validation protocol. All models achieved excellent performance, with cross-validated F1-scores exceeding 0.97 and balanced accuracy up to 0.993. The best results were obtained for ResNet50 and ConvNeXt Tiny (F1-score: 0.9879 ± 0.0077 and 0.9860 ± 0.0073, balanced accuracy: 0.9922 ± 0.0054 and 0.9927 ± 0.0042, respectively), indicating outstanding sensitivity and specificity for both flowering and non-flowering classes. Classical CNNs (VGG16, ResNet50, and ConvNeXt Tiny) demonstrated slightly superior robustness compared to transformer-based models, though all architectures maintained high generalization and minimal variance across folds. The integrated quality assessment and classification pipeline enables scalable, high-throughput monitoring of flowering phases in natural environments. The proposed methodology is adaptable to other plant species and locations, supporting future ecological monitoring and climate studies. Our key contributions are as follows: (i) introducing an automated exposure-quality filtering stage for field imagery; (ii) publishing a curated, season-long dataset of Tilia cordata images; and (iii) providing the first systematic cross-validated benchmark that contrasts classical CNNs with transformer architectures for phenological phase recognition.</jats:p>
dc.description.accesstimeat_publication
dc.description.bibliographyil., bibliogr.
dc.description.financepublication_nocost
dc.description.financecost0,00
dc.description.if3,5
dc.description.number17
dc.description.points100
dc.description.versionfinal_published
dc.description.volume25
dc.identifier.doi10.3390/s25175326
dc.identifier.issn1424-8220
dc.identifier.urihttps://sciencerep.up.poznan.pl/handle/item/4505
dc.identifier.weblinkhttps://www.mdpi.com/1424-8220/25/17/5326
dc.languageen
dc.pbn.affiliationmechanical engineering
dc.relation.ispartofSensors
dc.relation.pagesart. 5326
dc.rightsCC-BY
dc.sciencecloudsend
dc.share.typeOPEN_JOURNAL
dc.subject.endeep learning
dc.subject.enconvolutional neural networks
dc.subject.envision transformer
dc.subject.enautomated image quality assessment
dc.subject.enTilia cordata
dc.subject.enflowering phase classification
dc.subject.enecological monitoring
dc.titleComparative Evaluation of CNN and Transformer Architectures for Flowering Phase Classification of Tilia cordata Mill. with Automated Image Quality Filtering
dc.title.volumeSpecial Issue Application of UAV and Sensing in Precision Agriculture
dc.typeJournalArticle
dspace.entity.typePublication
oaire.citation.issue17
oaire.citation.volume25