All Issue

2025 Vol.70, Issue 4 Preview Page

Original Research Article

Development of Rice Pre-harvest Sprouting Rate Measurement Program Using YOLOv8 Deep Learning Model

YOLOv8 딥러닝 모델을 이용한 벼 수발아율 측정 프로그램 개발

Soojin Jun1†
,
Younguk Kim2†
,
Myeongjin Kang3
,
Mi Hyun Cho4
,
Hyoja Oh3
,
HwangWeon Jeong5
,
Jae Il Lyu6
,
Jeongho Baek6
,
Seon-Hwa Bae7
,
Hongro Lee8
,
Hyeonso Ji9*
전 수진1†
,
김 영욱2†
,
강 명진3
,
조 미현4
,
오 효자3
,
정 황원5
,
유 재일6
,
백 정호6
,
배 선화7
,
이 홍로8
,
지 현소9*
1Researcher, Central-Northern Region Crop Research Center, National Institute of Crop Science, Suwon 16429, Korea
2Ph. D Student, Department of Computer Information Engineering, Kunsan National University, Gunsan 54150, Korea
3Research staff, Digital Breeding Convergence Division, National Institute of Agricultural Sciences, Jeonju 54874, Korea
4Research staff, Agri-Bio Supercomputing Center, National Institute of Agricultural Sciences, Jeonju 54874, Korea
5Researcher, Agri-Bio Supercomputing Center, National Institute of Agricultural Sciences, Jeonju 54874, Korea
6Researcher, Digital Breeding Convergence Division, National Institute of Horticultural and Herbal Science, Jeonju 54874, Korea
7Post-doctoral associate, Fruit Research Division, National Institute of Agricultural Sciences, Wanju 55365, Korea
8Professor, Department of Computer Information Engineering, Kunsan National University, Gunsan 54150, Korea
9Senior Researcher, Digital Breeding Convergence Division, National Institute of Agricultural Sciences, Jeonju 54874, Korea
1농촌진흥청 국립식량과학원 중북부작물연구센터 농업연구사
2군산대학교 컴퓨터정보공학과 박사과정
3농촌진흥청 국립농업과학원 디지털육종지원과 연구원
4농촌진흥청 국립농업과학원 슈퍼컴퓨팅센터 연구원
5농촌진흥청 국립농업과학원 슈퍼컴퓨팅센터 농업연구사
6농촌진흥청 국립농업과학원 디지털육종지원과 농업연구사
7농촌진흥청 국립원예특작과학원 과수기초기반과 박사후연구원
8군산대학교 컴퓨터정보공학과 교수
9농촌진흥청 국립농업과학원 디지털육종지원과 농업연구관
*Corresponding Author
†Theses authors contributed equally to this work.
1 December 2025. pp. 234-240
PDF XML Citation
Abstract

Pre-harvest Sprouting (PHS) in rice is a critical agronomic problem that significantly reduces grain yield and quality. Conventional PHS measurement relies on manual visual inspection, which is labor-intensive, time-consuming, and prone to assessor bias, rendering it unsuitable for high-throughput phenotyping. In this study, we developed a computer-based program for the automated measurement of PHS and seed germination rate. This program integrates a deep learning model for detecting germinated and ungerminated rice seeds and a user-friendly graphical user interface (GUI). A dataset of 132,200 seed images, produced under diverse background colors, was used for training the deep learning model. The optimized model achieved a mean average precision at IoU 0.5 (mAP50) of 0.987 and an mAP50-90 of 0.916, demonstrating a robust performance under varying imaging conditions. The developed GUI facilitates key functions, including image loading, single and batch detection, visualization of detection results, automatic calculation of PHS and germination rates, and exporting of data. To assess the reliability of the method, PHS rates measured using threshed rice seeds were compared against those of conventional measurement methods using unthreshed rice seeds across 20 rice cultivars. A strong positive correlation (R2 = 0.8909) demonstrated that the proposed image analysis approach using threshed seeds serves as a reliable alternative to the conventional manual method using unthreshed rice seeds. The program developed in this study is convenient to use and significantly enhances the reproducibility and efficiency of rice PHS and germination rate measurement, thereby enabling streamlined phenotyping for rice breeding programs and other rice research.

Keywords
deep learning
pre-harvest sprouting
rice
YOLO
References
1

Cobb, J. N., G. DeClerck, A. Greenberg, R. Clark, and S. McCouch. 2013. Next-generation phenotyping: requirements and strategies for enhancing our understanding of genotype- phenotype relationships and its relevance to crop improvement. Theor Appl Genet 126(4):867-887.

10.1007/s00122-013-2066-023471459PMC3607725
2

Colmer, J., C. M. O'Neill, R. Wells, A. Bostrom, D. Reynolds, D. Websdale, G. Shiralagi, W. Lu, Q. Lou, T. Le Cornu, J. Ball, J. Renema, G. Flores Andaluz, R. Benjamins, S. Penfield, and J. Zhou. 2020. SeedGerm: a cost-effective phenotyping platform for automated seed imaging and machine-learning based phenotypic analysis of crop seed germination. New Phytol. 228(2) : 778-793.

10.1111/nph.16736
3

Diwan, T., G. Anirudh, and J. V. Tembhurne. 2023. Object detection using YOLO: challenges, architectural successors, datasets and applications. Multimedia Tools and Applications. 82(6) : 9243-9275.

10.1007/s11042-022-13644-y35968414PMC9358372
4

Dutta, A. and A. Zisserman. 2019. The VIA Annotation Software for Images, Audio and Video. Paper presented at the 27th ACM International Conference on Multimedia; Nice, France.

10.1145/3343031.3350535
5

Genze, N., R. Bharti, M. Grieb, S. J. Schultheiss, and D. G. Grimm. 2020. Accurate machine learning-based germination detection, prediction and quality assessment of three grain crops. Plant Methods. 16(1) : 157.

10.1186/s13007-020-00699-x33353559PMC7754596
6

Jeon, D., Y. Kang, S. Lee, S. Choi, Y. Sung, T.-H. Lee, and C. Kim. 2023. Digitalizing breeding in plants: A new trend of next-generation breeding based on genomic prediction. Front. Plant Sci. 14 : 1092584.

10.3389/fpls.2023.109258436743488PMC9892199
7

Jun, S., M. H. Cho, H. Oh, Y. Kim, D. K. Yoon, M. Kang, H. Kim, S.-H. Bae, S. L. Kim, J. Baek, H. Jeong, J. I. Lyu, G.-S. Lee, C. Kim, and H. Ji. 2025. Mapping QTLs for PHS resistance and development of a deep learning model to measure PHS rate in japonica rice. The Plant Genome. 18(3) : e70109.

10.1002/tpg2.7010940884181PMC12397900
8

Lee, H., Y. Lee, W. Hwang, J. Jeong, S. Yang, C. Lee, and M. Choi. 2020. Investigation of changes in grain quality and physicochemical properties of rice according to the temperature during the ripening stage and preharvest sprouting. Korean J. Crop Sci. 65(4) : 294-302.

9

Li, Y., J. Xiao, L. Chen, X. Huang, Z. Cheng, B. Han, Q. Zhang, and C. Wu. 2018. Rice functional genomics research: past decade and future. Mol. Plant 11(3) : 359-380.

10.1016/j.molp.2018.01.007
10

Sohn, S.-I., S. Pandian, T. S. Kumar, Y. A. B. Zoclanclounon, P. Muthuramalingam, J. Shilpha, L. Satish, and M. Ramesh. 2021. Seed dormancy and pre-Harvest sprouting in rice-an updated overview. Int. J. Mol. Sci. 22 : 11804.

10.3390/ijms22211180434769234PMC8583970
11

Yao, Q., X. Zheng, G. Zhou, and J. Zhang. 2024. SGR-YOLO: a method for detecting seed germination rate in wild rice. Front. Plant Sci. 14 : 1305081.

10.3389/fpls.2023.130508138322421PMC10844399
12

Zhao, J., Y. Ma, K. Yong, M. Zhu, Y. Wang, Z. Luo, X. Wei, and X. Huang. 2023. Deep-learning-based automatic evaluation of rice seed germination rate. J. Sci. Food Agric. 103(4) : 1912-1924.

10.1002/jsfa.12318
Information
  • Publisher :The Korean Society of Crop Science
  • Publisher(Ko) :한국작물학회
  • Journal Title :The Korean Journal of Crop Science
  • Journal Title(Ko) :한국작물학회지
  • Volume : 70
  • No :4
  • Pages :234-240
  • Received Date : 2025-11-13
  • Revised Date : 2025-11-19
  • Accepted Date : 2025-11-19
  • DOI :https://doi.org/10.7740/kjcs.2025.70.4.234
Journal Informaiton The Korean Journal of Crop Science The Korean Journal of Crop Science
Online Submission
  • NRF
  • KOFST
  • KISTI Current Status
  • KISTI Cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close