Faster R-CNN-based deep learning for locating corn tassels in UAV imagery

Aziza Al-Zadjali; Yeyin Shi; Stephen Scott; Jitender S. Deogun; James Schnable

doi:10.1117/12.2560596

Faster R-CNN-based deep learning for locating corn tassels in UAV imagery

Aziza Al-Zadjali, Yeyin Shi, Stephen Scott, Jitender S. Deogun, James Schnable

Computer Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

1 Scopus citations

Abstract

Automating the detection of the corn tassels during owering time is important in corn breeding. To control pollination, after a tassel is visible, the plant should be checked daily for emerging ears. The conventional methods are labor-intensive and time-consuming. In this study, we developed a technique for automatic detecting and locating corn tassel in unmanned aerial vehicle (UAV) imagery with the state-of-the art Faster Region based Convolutional Neural Network (Faster R-CNN). Each raw image was divided into 1000 x 1000 pixels sub-images, and 2000 sub-images were manually annotated for tassel locations with bounding boxes as ground-truth data. 80% of the annotated sub-images were used as training data and the remaining 20% were used for testing. The performance of the trained Faster R-CNN model was evaluated by customized evaluation criteria. The model achieved good performance on tassel detection with mean average precision of 91.78% and F1 score up to 97.98%.

Original language	English (US)
Title of host publication	Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V
Editors	J. Alex Thomasson, Alfonso F. Torres-Rua
Publisher	SPIE
ISBN (Electronic)	9781510636057
DOIs	https://doi.org/10.1117/12.2560596
State	Published - 2020
Event	Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V 2020 - Virtual, Online, United States Duration: Apr 27 2020 → May 8 2020

Publication series

Name	Proceedings of SPIE - The International Society for Optical Engineering
Volume	11414
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V 2020
Country/Territory	United States
City	Virtual, Online
Period	4/27/20 → 5/8/20

Keywords

CNN
Faster R-CNN
Flowering
Object detection
Phenotyping

ASJC Scopus subject areas

Electronic, Optical and Magnetic Materials
Condensed Matter Physics
Computer Science Applications
Applied Mathematics
Electrical and Electronic Engineering

Access to Document

10.1117/12.2560596

Cite this

Al-Zadjali, A., Shi, Y., Scott, S., Deogun, J. S., & Schnable, J. (2020). Faster R-CNN-based deep learning for locating corn tassels in UAV imagery. In J. A. Thomasson, & A. F. Torres-Rua (Eds.), Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V Article 1141406 (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11414). SPIE. https://doi.org/10.1117/12.2560596

Faster R-CNN-based deep learning for locating corn tassels in UAV imagery. / Al-Zadjali, Aziza; Shi, Yeyin; Scott, Stephen et al.
Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V. ed. / J. Alex Thomasson; Alfonso F. Torres-Rua. SPIE, 2020. 1141406 (Proceedings of SPIE - The International Society for Optical Engineering; Vol. 11414).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Al-Zadjali, A, Shi, Y, Scott, S, Deogun, JS & Schnable, J 2020, Faster R-CNN-based deep learning for locating corn tassels in UAV imagery. in JA Thomasson & AF Torres-Rua (eds), Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V., 1141406, Proceedings of SPIE - The International Society for Optical Engineering, vol. 11414, SPIE, Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V 2020, Virtual, Online, United States, 4/27/20. https://doi.org/10.1117/12.2560596

Al-Zadjali A, Shi Y, Scott S, Deogun JS, Schnable J. Faster R-CNN-based deep learning for locating corn tassels in UAV imagery. In Thomasson JA, Torres-Rua AF, editors, Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V. SPIE. 2020. 1141406. (Proceedings of SPIE - The International Society for Optical Engineering). doi: 10.1117/12.2560596

@inproceedings{6065d55d663c4b25a725c04b96413ff5,

title = "Faster R-CNN-based deep learning for locating corn tassels in UAV imagery",

abstract = "Automating the detection of the corn tassels during owering time is important in corn breeding. To control pollination, after a tassel is visible, the plant should be checked daily for emerging ears. The conventional methods are labor-intensive and time-consuming. In this study, we developed a technique for automatic detecting and locating corn tassel in unmanned aerial vehicle (UAV) imagery with the state-of-the art Faster Region based Convolutional Neural Network (Faster R-CNN). Each raw image was divided into 1000 x 1000 pixels sub-images, and 2000 sub-images were manually annotated for tassel locations with bounding boxes as ground-truth data. 80% of the annotated sub-images were used as training data and the remaining 20% were used for testing. The performance of the trained Faster R-CNN model was evaluated by customized evaluation criteria. The model achieved good performance on tassel detection with mean average precision of 91.78% and F1 score up to 97.98%.",

keywords = "CNN, Faster R-CNN, Flowering, Object detection, Phenotyping",

author = "Aziza Al-Zadjali and Yeyin Shi and Stephen Scott and Deogun, {Jitender S.} and James Schnable",

note = "Funding Information: This research was funded by the Seed Grant of the University of Nebraska Collaboration Initiative (A-0000000325), the Nebraska Corn Board, and the Nebraska Agricultural Experiment Station through the Hatch Act capacity funding program (Accession Number 1011130) from the USDA National Institute of Food and Agriculture. Publisher Copyright: {\textcopyright} 2020 SPIE.; Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V 2020 ; Conference date: 27-04-2020 Through 08-05-2020",

year = "2020",

doi = "10.1117/12.2560596",

language = "English (US)",

series = "Proceedings of SPIE - The International Society for Optical Engineering",

publisher = "SPIE",

editor = "Thomasson, {J. Alex} and Torres-Rua, {Alfonso F.}",

booktitle = "Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V",

}

TY - GEN

T1 - Faster R-CNN-based deep learning for locating corn tassels in UAV imagery

AU - Al-Zadjali, Aziza

AU - Shi, Yeyin

AU - Scott, Stephen

AU - Deogun, Jitender S.

AU - Schnable, James

N1 - Funding Information: This research was funded by the Seed Grant of the University of Nebraska Collaboration Initiative (A-0000000325), the Nebraska Corn Board, and the Nebraska Agricultural Experiment Station through the Hatch Act capacity funding program (Accession Number 1011130) from the USDA National Institute of Food and Agriculture. Publisher Copyright: © 2020 SPIE.

PY - 2020

Y1 - 2020

N2 - Automating the detection of the corn tassels during owering time is important in corn breeding. To control pollination, after a tassel is visible, the plant should be checked daily for emerging ears. The conventional methods are labor-intensive and time-consuming. In this study, we developed a technique for automatic detecting and locating corn tassel in unmanned aerial vehicle (UAV) imagery with the state-of-the art Faster Region based Convolutional Neural Network (Faster R-CNN). Each raw image was divided into 1000 x 1000 pixels sub-images, and 2000 sub-images were manually annotated for tassel locations with bounding boxes as ground-truth data. 80% of the annotated sub-images were used as training data and the remaining 20% were used for testing. The performance of the trained Faster R-CNN model was evaluated by customized evaluation criteria. The model achieved good performance on tassel detection with mean average precision of 91.78% and F1 score up to 97.98%.

AB - Automating the detection of the corn tassels during owering time is important in corn breeding. To control pollination, after a tassel is visible, the plant should be checked daily for emerging ears. The conventional methods are labor-intensive and time-consuming. In this study, we developed a technique for automatic detecting and locating corn tassel in unmanned aerial vehicle (UAV) imagery with the state-of-the art Faster Region based Convolutional Neural Network (Faster R-CNN). Each raw image was divided into 1000 x 1000 pixels sub-images, and 2000 sub-images were manually annotated for tassel locations with bounding boxes as ground-truth data. 80% of the annotated sub-images were used as training data and the remaining 20% were used for testing. The performance of the trained Faster R-CNN model was evaluated by customized evaluation criteria. The model achieved good performance on tassel detection with mean average precision of 91.78% and F1 score up to 97.98%.

KW - CNN

KW - Faster R-CNN

KW - Flowering

KW - Object detection

KW - Phenotyping

UR - http://www.scopus.com/inward/record.url?scp=85089092044&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85089092044&partnerID=8YFLogxK

U2 - 10.1117/12.2560596

DO - 10.1117/12.2560596

M3 - Conference contribution

AN - SCOPUS:85089092044

T3 - Proceedings of SPIE - The International Society for Optical Engineering

BT - Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V

A2 - Thomasson, J. Alex

A2 - Torres-Rua, Alfonso F.

PB - SPIE

T2 - Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping V 2020

Y2 - 27 April 2020 through 8 May 2020

ER -

Faster R-CNN-based deep learning for locating corn tassels in UAV imagery

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this