Towards Automated Analysis of Grain Spikes in Greenhouse Images
Using Neural Network Approaches: A Comparative Investigation of
Six Methods

J 2021

Towards Automated Analysis of Grain Spikes in Greenhouse Images Using Neural Network Approaches: A Comparative Investigation of Six Methods

ULLAH, Sajid, Michael HENKE, N. NARISETTI, K. PANZAROVA, M. TRTILEK et. al.

Základní údaje

Originální název

Towards Automated Analysis of Grain Spikes in Greenhouse Images Using Neural Network Approaches: A Comparative Investigation of Six Methods

Autoři

ULLAH, Sajid (586 Pákistán, domácí), Michael HENKE (276 Německo, domácí), N. NARISETTI, K. PANZAROVA, M. TRTILEK, Jan HEJÁTKO (203 Česká republika, garant, domácí) a E. GLADILIN

Vydání

IEEE Sensors Journal, BASEL, IEEE Sensors Council, 2021, 1424-8220

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10406 Analytical chemistry

Stát vydavatele

Švýcarsko

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 3.847

Kód RIV

RIV/00216224:14740/21:00124192

Organizační jednotka

Středoevropský technologický institut

DOI

http://dx.doi.org/10.3390/s21227441

UT WoS

000726906200001

Klíčová slova anglicky

high-throughput plant image analysis; spike detection; spike segmentation; deep learning; automated plant phenotyping

Štítky

rivok

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 18. 5. 2022 13:35, Mgr. Pavla Foltynová, Ph.D.

Anotace

V originále

Automated analysis of small and optically variable plant organs, such as grain spikes, is highly demanded in quantitative plant science and breeding. Previous works primarily focused on the detection of prominently visible spikes emerging on the top of the grain plants growing in field conditions. However, accurate and automated analysis of all fully and partially visible spikes in greenhouse images renders a more challenging task, which was rarely addressed in the past. A particular difficulty for image analysis is represented by leaf-covered, occluded but also matured spikes of bushy crop cultivars that can hardly be differentiated from the remaining plant biomass. To address the challenge of automated analysis of arbitrary spike phenotypes in different grain crops and optical setups, here, we performed a comparative investigation of six neural network methods for pattern detection and segmentation in RGB images, including five deep and one shallow neural network. Our experimental results demonstrate that advanced deep learning methods show superior performance, achieving over 90% accuracy by detection and segmentation of spikes in wheat, barley and rye images. However, spike detection in new crop phenotypes can be performed more accurately than segmentation. Furthermore, the detection and segmentation of matured, partially visible and occluded spikes, for which phenotypes substantially deviate from the training set of regular spikes, still represent a challenge to neural network models trained on a limited set of a few hundreds of manually labeled ground truth images. Limitations and further potential improvements of the presented algorithmic frameworks for spike image analysis are discussed. Besides theoretical and experimental investigations, we provide a GUI-based tool (SpikeApp), which shows the application of pre-trained neural networks to fully automate spike detection, segmentation and phenotyping in images of greenhouse-grown plants.

Návaznosti

EF16_026/0008446, projekt VaV

Název: Integrace signálu a epigenetické reprogramování pro produktivitu rostlin

Citovat

ULLAH, Sajid, Michael HENKE, N. NARISETTI, K. PANZAROVA, M. TRTILEK, Jan HEJÁTKO a E. GLADILIN. Towards Automated Analysis of Grain Spikes in Greenhouse Images Using Neural Network Approaches: A Comparative Investigation of Six Methods. IEEE Sensors Journal. BASEL: IEEE Sensors Council, 2021, roč. 21, č. 22, s. „7441“, 22 s. ISSN 1424-8220. Dostupné z: https://dx.doi.org/10.3390/s21227441.

@article{1834769,
   author = {Ullah, Sajid and Henke, Michael and Narisetti, N. and Panzarova, K. and Trtilek, M. and Hejátko, Jan and Gladilin, E.},
   article_location = {BASEL},
   article_number = {22},
   doi = {http://dx.doi.org/10.3390/s21227441},
   keywords = {high-throughput plant image analysis; spike detection; spike segmentation; deep learning; automated plant phenotyping},
   language = {eng},
   issn = {1424-8220},
   journal = {IEEE Sensors Journal},
   title = {Towards Automated Analysis of Grain Spikes in Greenhouse Images Using Neural Network Approaches: A Comparative Investigation of Six Methods},
   url = {https://www.mdpi.com/1424-8220/21/22/7441},
   volume = {21},
   year = {2021}
}

TY  - JOUR
ID  - 1834769
AU  - Ullah, Sajid - Henke, Michael - Narisetti, N. - Panzarova, K. - Trtilek, M. - Hejátko, Jan - Gladilin, E.
PY  - 2021
TI  - Towards Automated Analysis of Grain Spikes in Greenhouse Images Using Neural Network Approaches: A Comparative Investigation of Six Methods
JF  - IEEE Sensors Journal
VL  - 21
IS  - 22
SP  - „7441“
EP  - „7441“
PB  - IEEE Sensors Council
SN  - 14248220
KW  - high-throughput plant image analysis
KW  - spike detection
KW  - spike segmentation
KW  - deep learning
KW  - automated plant phenotyping
UR  - https://www.mdpi.com/1424-8220/21/22/7441
N2  - Automated analysis of small and optically variable plant organs, such as grain spikes, is highly demanded in quantitative plant science and breeding. Previous works primarily focused on the detection of prominently visible spikes emerging on the top of the grain plants growing in field conditions. However, accurate and automated analysis of all fully and partially visible spikes in greenhouse images renders a more challenging task, which was rarely addressed in the past. A particular difficulty for image analysis is represented by leaf-covered, occluded but also matured spikes of bushy crop cultivars that can hardly be differentiated from the remaining plant biomass. To address the challenge of automated analysis of arbitrary spike phenotypes in different grain crops and optical setups, here, we performed a comparative investigation of six neural network methods for pattern detection and segmentation in RGB images, including five deep and one shallow neural network. Our experimental results demonstrate that advanced deep learning methods show superior performance, achieving over 90% accuracy by detection and segmentation of spikes in wheat, barley and rye images. However, spike detection in new crop phenotypes can be performed more accurately than segmentation. Furthermore, the detection and segmentation of matured, partially visible and occluded spikes, for which phenotypes substantially deviate from the training set of regular spikes, still represent a challenge to neural network models trained on a limited set of a few hundreds of manually labeled ground truth images. Limitations and further potential improvements of the presented algorithmic frameworks for spike image analysis are discussed. Besides theoretical and experimental investigations, we provide a GUI-based tool (SpikeApp), which shows the application of pre-trained neural networks to fully automate spike detection, segmentation and phenotyping in images of greenhouse-grown plants.
ER  -

ULLAH, Sajid, Michael HENKE, N. NARISETTI, K. PANZAROVA, M. TRTILEK, Jan HEJÁTKO a E. GLADILIN. Towards Automated Analysis of Grain Spikes in Greenhouse Images Using Neural Network Approaches: A Comparative Investigation of Six Methods. \textit{IEEE Sensors Journal}. BASEL: IEEE Sensors Council, 2021, roč.~21, č.~22, s.~„7441“, 22 s. ISSN~1424-8220. Dostupné z: https://dx.doi.org/10.3390/s21227441.

Podrobný výpis o publikaci