Semantic segmentation of plant roots from RGB (mini-) rhizotron images-generalisation potential and false positives of established methods and advanced deep-learning models

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F62156489%3A43410%2F23%3A43924191" target="_blank" >RIV/62156489:43410/23:43924191 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1186/s13007-023-01101-2" target="_blank" >https://doi.org/10.1186/s13007-023-01101-2</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1186/s13007-023-01101-2" target="_blank" >10.1186/s13007-023-01101-2</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Semantic segmentation of plant roots from RGB (mini-) rhizotron images-generalisation potential and false positives of established methods and advanced deep-learning models

Popis výsledku v původním jazyce

Background Manual analysis of (mini-)rhizotron (MR) images is tedious. Several methods have been proposed for semantic root segmentation based on homogeneous, single-source MR datasets. Recent advances in deep learning (DL) have enabled automated feature extraction, but comparisons of segmentation accuracy, false positives and transferability are virtually lacking. Here we compare six state-of-the-art methods and propose two improved DL models for semantic root segmentation using a large MR dataset with and without augmented data. We determine the performance of the methods on a homogeneous maize dataset, and a mixed dataset of &gt; 8 species (mixtures), 6 soil types and 4 imaging systems. The generalisation potential of the derived DL models is determined on a distinct, unseen dataset. Results The best performance was achieved by the U-Net models; the more complex the encoder the better the accuracy and generalisation of the model. The heterogeneous mixed MR dataset was a particularly challenging for the non-U-Net techniques. Data augmentation enhanced model performance. We demonstrated the improved performance of deep meta-architectures and feature extractors, and a reduction in the number of false positives. Conclusions Although correction factors are still required to match human labelled root lengths, neural network architectures greatly reduce the time required to compute the root length. The more complex architectures illustrate how future improvements in root segmentation within MR images can be achieved, particularly reaching higher segmentation accuracies and model generalisation when analysing real-world datasets with artefacts-limiting the need for model retraining.

Název v anglickém jazyce

Semantic segmentation of plant roots from RGB (mini-) rhizotron images-generalisation potential and false positives of established methods and advanced deep-learning models

Popis výsledku anglicky

Background Manual analysis of (mini-)rhizotron (MR) images is tedious. Several methods have been proposed for semantic root segmentation based on homogeneous, single-source MR datasets. Recent advances in deep learning (DL) have enabled automated feature extraction, but comparisons of segmentation accuracy, false positives and transferability are virtually lacking. Here we compare six state-of-the-art methods and propose two improved DL models for semantic root segmentation using a large MR dataset with and without augmented data. We determine the performance of the methods on a homogeneous maize dataset, and a mixed dataset of &gt; 8 species (mixtures), 6 soil types and 4 imaging systems. The generalisation potential of the derived DL models is determined on a distinct, unseen dataset. Results The best performance was achieved by the U-Net models; the more complex the encoder the better the accuracy and generalisation of the model. The heterogeneous mixed MR dataset was a particularly challenging for the non-U-Net techniques. Data augmentation enhanced model performance. We demonstrated the improved performance of deep meta-architectures and feature extractors, and a reduction in the number of false positives. Conclusions Although correction factors are still required to match human labelled root lengths, neural network architectures greatly reduce the time required to compute the root length. The more complex architectures illustrate how future improvements in root segmentation within MR images can be achieved, particularly reaching higher segmentation accuracies and model generalisation when analysing real-world datasets with artefacts-limiting the need for model retraining.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

40102 - Forestry

Projekt

—

Návaznosti

O - Projekt operacniho programu

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Plant Methods

ISSN

1746-4811

e-ISSN

1746-4811

Svazek periodika

19

Číslo periodika v rámci svazku

6 November

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

15

Strana od-do

122

Kód UT WoS článku

001099034900001

EID výsledku v databázi Scopus

2-s2.0-85175806863