Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation

Abstract Premise Deep learning has become increasingly important in the analysis of digitized herbarium collections, which comprise millions of scans that provide valuable resources for studying plant evolution and biodiversity. However, leveraging deep learning algorithms to analyze these scans pre...

Full description

Saved in:
Bibliographic Details
Main Authors: Hanane Ariouat, Youcef Sklab, Edi Prifti, Jean‐Daniel Zucker, Eric Chenin
Format: Article
Language:English
Published: Wiley 2025-03-01
Series:Applications in Plant Sciences
Subjects:
Online Access:https://doi.org/10.1002/aps3.70000
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849311430568837120
author Hanane Ariouat
Youcef Sklab
Edi Prifti
Jean‐Daniel Zucker
Eric Chenin
author_facet Hanane Ariouat
Youcef Sklab
Edi Prifti
Jean‐Daniel Zucker
Eric Chenin
author_sort Hanane Ariouat
collection DOAJ
description Abstract Premise Deep learning has become increasingly important in the analysis of digitized herbarium collections, which comprise millions of scans that provide valuable resources for studying plant evolution and biodiversity. However, leveraging deep learning algorithms to analyze these scans presents significant challenges, partly due to the heterogeneous nature of the non‐plant material that forms the background of the scans. We hypothesize that removing such backgrounds can improve the performance of these algorithms. Methods We propose a novel method based on deep learning to segment and generate plant masks from herbarium scans and subsequently remove the non‐plant backgrounds. The semi‐automatic preprocessing stages involve the identification and removal of non‐plant elements, substantially reducing the manual effort required to prepare the training dataset. Results The results highlight the importance of effective image segmentation, which achieved an F1 score of up to 96.6%. Moreover, when used in classification models for plant morphological trait identification, the images resulting from segmentation improved classification accuracy by up to 3% and F1 score by up to 7% compared to non‐segmented images. Discussion Our approach isolates plant elements in herbarium scans by removing background elements to improve classification tasks. We demonstrate that image segmentation significantly enhances the performance of plant morphological trait identification models.
format Article
id doaj-art-f9fe3e01b8ec400eba83efc847df53b3
institution Kabale University
issn 2168-0450
language English
publishDate 2025-03-01
publisher Wiley
record_format Article
series Applications in Plant Sciences
spelling doaj-art-f9fe3e01b8ec400eba83efc847df53b32025-08-20T03:53:23ZengWileyApplications in Plant Sciences2168-04502025-03-01132n/an/a10.1002/aps3.70000Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentationHanane Ariouat0Youcef Sklab1Edi Prifti2Jean‐Daniel Zucker3Eric Chenin4Institut de Recherche pour le Développement (IRD) Sorbonne Université UMMISCO, F‐93143, Bondy FranceInstitut de Recherche pour le Développement (IRD) Sorbonne Université UMMISCO, F‐93143, Bondy FranceInstitut de Recherche pour le Développement (IRD) Sorbonne Université UMMISCO, F‐93143, Bondy FranceInstitut de Recherche pour le Développement (IRD) Sorbonne Université UMMISCO, F‐93143, Bondy FranceInstitut de Recherche pour le Développement (IRD) Sorbonne Université UMMISCO, F‐93143, Bondy FranceAbstract Premise Deep learning has become increasingly important in the analysis of digitized herbarium collections, which comprise millions of scans that provide valuable resources for studying plant evolution and biodiversity. However, leveraging deep learning algorithms to analyze these scans presents significant challenges, partly due to the heterogeneous nature of the non‐plant material that forms the background of the scans. We hypothesize that removing such backgrounds can improve the performance of these algorithms. Methods We propose a novel method based on deep learning to segment and generate plant masks from herbarium scans and subsequently remove the non‐plant backgrounds. The semi‐automatic preprocessing stages involve the identification and removal of non‐plant elements, substantially reducing the manual effort required to prepare the training dataset. Results The results highlight the importance of effective image segmentation, which achieved an F1 score of up to 96.6%. Moreover, when used in classification models for plant morphological trait identification, the images resulting from segmentation improved classification accuracy by up to 3% and F1 score by up to 7% compared to non‐segmented images. Discussion Our approach isolates plant elements in herbarium scans by removing background elements to improve classification tasks. We demonstrate that image segmentation significantly enhances the performance of plant morphological trait identification models.https://doi.org/10.1002/aps3.70000deep learningherbarium scanssemantic segmentationtrait classification
spellingShingle Hanane Ariouat
Youcef Sklab
Edi Prifti
Jean‐Daniel Zucker
Eric Chenin
Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation
Applications in Plant Sciences
deep learning
herbarium scans
semantic segmentation
trait classification
title Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation
title_full Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation
title_fullStr Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation
title_full_unstemmed Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation
title_short Enhancing plant morphological trait identification in herbarium collections through deep learning–based segmentation
title_sort enhancing plant morphological trait identification in herbarium collections through deep learning based segmentation
topic deep learning
herbarium scans
semantic segmentation
trait classification
url https://doi.org/10.1002/aps3.70000
work_keys_str_mv AT hananeariouat enhancingplantmorphologicaltraitidentificationinherbariumcollectionsthroughdeeplearningbasedsegmentation
AT youcefsklab enhancingplantmorphologicaltraitidentificationinherbariumcollectionsthroughdeeplearningbasedsegmentation
AT ediprifti enhancingplantmorphologicaltraitidentificationinherbariumcollectionsthroughdeeplearningbasedsegmentation
AT jeandanielzucker enhancingplantmorphologicaltraitidentificationinherbariumcollectionsthroughdeeplearningbasedsegmentation
AT ericchenin enhancingplantmorphologicaltraitidentificationinherbariumcollectionsthroughdeeplearningbasedsegmentation