High-throughput spike detection in greenhouse cultivated grain crops with attention mechanisms based deep learning models

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14310%2F24%3A00135605" target="_blank" >RIV/00216224:14310/24:00135605 - isvavai.cz</a>

Výsledek na webu

<a href="https://spj.science.org/doi/10.34133/plantphenomics.0155" target="_blank" >https://spj.science.org/doi/10.34133/plantphenomics.0155</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.34133/plantphenomics.0155" target="_blank" >10.34133/plantphenomics.0155</a>

Jazyk výsledku

angličtina

Název v původním jazyce

High-throughput spike detection in greenhouse cultivated grain crops with attention mechanisms based deep learning models

Popis výsledku v původním jazyce

Detection of spikes is the first important step towards image-based quantitative assessment of crop yield. However, spikes of grain plants occupy only a tiny fraction of the image area and often emerge in the middle of the mass of plant leaves that exhibit similar colors as spike regions. Consequently, accurate detection of grain spikes renders, in general, a non-trivial task even for advanced, state-of-the-art deep learning neural networks (DNNs). To improve pattern detection in spikes, we propose architectural changes to Faster-RCNN (FRCNN) by reducing feature extraction layers and introducing a global attention module. The performance of our extended FRCNN-A vs. conventional FRCNN was compared on images of different European wheat cultivars, including ’difficult’ bushy phenotypes from two different phenotyping facilities and optical setups. Our experimental results show that introduced architectural adaptations in FRCNN-A helped to improve spike detection accuracy in inner regions. The mAP of FRCNN and FRCNN-A on inner spikes is 76.0% and 81.0%, respectively, while on the state-of-the-art detection DNNs, Swin Transformer mAP is 83.0%. As a lightweight network, FRCNN-A is faster than FRCNN and Swin Transformer on both baseline and augmented training datasets. On the FastGAN augmented dataset, FRCNN achieved mAP of 84.24%, FRCNN-A 85.0%, and the Swin Transformer 89.45%. The increase in mAP of DNNs on the augmented datasets is proportional to the amount of the IPK original and augmented images. Overall, this study indicates a superior performance of attention mechanisms-based deep learning models in detecting small and subtle features of grain spikes.

Název v anglickém jazyce

High-throughput spike detection in greenhouse cultivated grain crops with attention mechanisms based deep learning models

Popis výsledku anglicky

Detection of spikes is the first important step towards image-based quantitative assessment of crop yield. However, spikes of grain plants occupy only a tiny fraction of the image area and often emerge in the middle of the mass of plant leaves that exhibit similar colors as spike regions. Consequently, accurate detection of grain spikes renders, in general, a non-trivial task even for advanced, state-of-the-art deep learning neural networks (DNNs). To improve pattern detection in spikes, we propose architectural changes to Faster-RCNN (FRCNN) by reducing feature extraction layers and introducing a global attention module. The performance of our extended FRCNN-A vs. conventional FRCNN was compared on images of different European wheat cultivars, including ’difficult’ bushy phenotypes from two different phenotyping facilities and optical setups. Our experimental results show that introduced architectural adaptations in FRCNN-A helped to improve spike detection accuracy in inner regions. The mAP of FRCNN and FRCNN-A on inner spikes is 76.0% and 81.0%, respectively, while on the state-of-the-art detection DNNs, Swin Transformer mAP is 83.0%. As a lightweight network, FRCNN-A is faster than FRCNN and Swin Transformer on both baseline and augmented training datasets. On the FastGAN augmented dataset, FRCNN achieved mAP of 84.24%, FRCNN-A 85.0%, and the Swin Transformer 89.45%. The increase in mAP of DNNs on the augmented datasets is proportional to the amount of the IPK original and augmented images. Overall, this study indicates a superior performance of attention mechanisms-based deep learning models in detecting small and subtle features of grain spikes.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/EF16_026%2F0008446" target="_blank" >EF16_026/0008446: Integrace signálu a epigenetické reprogramování pro produktivitu rostlin</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2024

Kód důvěrnosti údajů

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

Název periodika

Plant Phenomics

ISSN

2643-6515

e-ISSN

2643-6515

Svazek periodika

6

Číslo periodika v rámci svazku

March

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

11

Strana od-do

1-11

Kód UT WoS článku

001231155500001

EID výsledku v databázi Scopus

2-s2.0-85191439858