High-precision non-destructive identification method for maize seed varieties based on fusion of LED white-light imaging and YOLOv8 model

GENG Lijie; HOU Lingqi; FU Zhenxiang; BI Chenglong; ZHU Huajie; LIU Xuanyi; HOU Xingsong; YANG Kun; LI Jin

doi:10.37188/CJLCD.2026-0072

您当前的位置：

首页 >

文章列表页 >

High-precision non-destructive identification method for maize seed varieties based on fusion of LED white-light imaging and YOLOv8 model

Image Processing | 更新时间：2026-09-12

- High-precision non-destructive identification method for maize seed varieties based on fusion of LED white-light imaging and YOLOv8 model
- Chinese Journal of Liquid Crystals and Displays Vol. 41, Issue 6, Pages: 866-877(2026)
- 作者机构：
  
  1.郑州轻工业大学电子信息学院河南省光电传感与信息处理工程技术研究中心，河南郑州 450000
  2.河南九圣禾新科种业有限公司，河南新乡 453000
  3.中原工学院物理与光电工程学院，河南郑州 450007
  4.西北农林科技大学机械与电子工程学院，陕西咸阳 712100
- 作者简介：
- 基金信息：
  
  the Key Research and Development Projects of Henan Province(241111212600);Training Plan of Young Backbone Teachers in Universities of Henan Province(2023GGJS088);Henan Provincial Science and Technology Research Project(242102210157;252102210188);Shaanxi Provincial Qinchuangyuan Innovation and Entrepreneurship Talent Project(QCYRCXM-2023-111)
- DOI：10.37188/CJLCD.2026-0072
  CLC： TP391.41
- CSTR：32172.14.CJLCD.2026-0072
- Received：27 April 2026，
  
  Revised：2026-05-16，
  
  Published：06 June 2026
- 稿件说明：
移动端阅览
GENG Lijie, HOU Lingqi, FU Zhenxiang, et al. High-precision non-destructive identification method for maize seed varieties based on fusion of LED white-light imaging and YOLOv8 model[J]. Chinese Journal of Liquid Crystals and Displays, 2026, 41(6): 866-877.
DOI：

GENG Lijie, HOU Lingqi, FU Zhenxiang, et al. High-precision non-destructive identification method for maize seed varieties based on fusion of LED white-light imaging and YOLOv8 model[J]. Chinese Journal of Liquid Crystals and Displays, 2026, 41(6): 866-877. DOI： 10.37188/CJLCD.2026-0072. CSTR： 32172.14.CJLCD.2026-0072.

摘要

种子纯度是决定农民收成与收入的关键因素，而快速精准的品种识别技术则是保障种子纯度的核心。本研究提出基于LED白光成像与YOLOv8模型融合的玉米种子无损检测方法，实验证实该方法兼具快

速、高效、高精度和无损等优势。通过搭建优化的LED白光反射成像系统，完成了4个玉米品种共12 008粒种子的图像采集与标注。对比实验显示，YOLOv8m模型的各项核心指标显著优于轻量级YOLOv8s，有效解决了形态相似品种识别精度不足的难题。该模型对4个品种的识别平均精度（AP）、召回率（Recall，

）和F1分数均达99.8%，mAP@0.5值高达99.4%。实时性测试中，YOLOv8m单图处理耗时68.6 ms，满足种子分选系统的低延迟需求。研究还揭示了数据集规模对模型性能的影响规律，证实YOLOv8m在不同规模数据集上均保持卓越的稳定性。本研究为种子品种识别领域提供了一种精确又实用的技术方案。

Abstract

Seed purity is a critical factor determining farmers’ harvests and income， and rapid， accurate variety identification technology is the core to ensuring purity. This study proposes a non-destructive detection method for maize seeds based on the fusion of LED white-light imaging and the YOLOv8 model. Experiments confirm that this method combines the advantages of fast speed， high efficiency， high accuracy， and non-destructiveness. By constructing an optimized LED white-light reflection imaging system， we completed image acquisition and annotation for a total of 12 008 seeds from four maize varieties. Comparative experiments show that the YOLOv8m model significantly outperforms the lightweight YOLOv8s in various core metrics， effectively addressing the issue of insufficient identification accuracy for morphologically similar varieties. The model achieved an average precision（AP）， recall rate（

）， and F1 score of 99.8% for the four varieties， with an mAP@0.5 value as high as 99.4%. In real-time performance tests， the YOLOv8m model took 68.6 ms to process a single image， fully meeting the low-latency requirements of seed sorting systems. The research also reveals the impact pattern of dataset scale on model performance， confirming that YOLOv8m maintains excellent stability across datasets of different scales. The proposed method provides a precise and practical technical solution for the field of seed variety identification.

关键词

Keywords

references

GARCÍA-LARA S ， SERNA-SALDIVAR S O . Corn history and culture ［M］//SERNA-SALDIVAR S O. Corn . 3rd ed . Cambridge ： Woodhead Publishing and AACC International Press ， 2019 ： 1 - 18 . doi: 10.1016/b978-0-12-811971-6.00001-2 http://dx.doi.org/10.1016/b978-0-12-811971-6.00001-2

孙建华，邹原东 . 关于玉米种子纯度鉴定方法分析［J］. 农业科学， 2020 ， 10 （ 5 ）： 292 - 298 . doi: 10.12677/hjas.2020.105043 http://dx.doi.org/10.12677/hjas.2020.105043

SUN J H ， ZOU Y D . Analysis on the method of corn seed purity identification ［J］. Hans J. Agric. Sci. ， 2020 ， 10 （ 5 ）： 292 - 298 . （in Chinese） . doi: 10.12677/hjas.2020.105043 http://dx.doi.org/10.12677/hjas.2020.105043

PARK H S ， CHOI K C ， KIM J H ， et al . Discrimination and quantification between annual ryegrass and perennial ryegrass seeds by near-infrared spectroscopy ［J］. J. Anim. Plant Sci. ， 2016 ， 26 （ 5 ）： 1278 - 1283 .

SUNDARAM R M ， NAVEENKUMAR B ， BIRADAR S K ， et al . Identification of informative SSR markers capable of distinguishing hybrid rice parental lines and their utilization in seed purity assessment ［J］. Euphytica ， 2008 ， 163 （ 2 ）： 215 - 224 . doi: 10.1007/s10681-007-9630-0 http://dx.doi.org/10.1007/s10681-007-9630-0

SATTURU V ， RANI D ， GATTU S ， et al . DNA fingerprinting for identification of rice varieties and seed genetic purity assessment ［J］. Agric. Res. ， 2018 ， 7 （ 4 ）： 379 - 390 . doi: 10.1007/s40003-018-0324-8 http://dx.doi.org/10.1007/s40003-018-0324-8

PALLAVI H M ， GOWDA R ， SHADAKSHARI Y G ， et al . Identification of SSR markers for hybridity and seed genetic purity testing in sunflower （ Helianthus annuus L.）［J］. Helia ， 2011 ， 34 （ 54 ）： 59 - 66 . doi: 10.2298/hel1154059p http://dx.doi.org/10.2298/hel1154059p

李浩光，李卫军，覃鸿，等 . 基于栈式自编码神经网络的包衣单籽粒玉米品种识别［J］. 农业机械学报， 2017 ， 48 （ S1 ）： 422 - 428 .

LI H G ， LI W J ， QIN H ， et al . Varietal identification for single maize seed based on stacked auto encoder neural network ［J］. Trans. Chin. Soc. Agric. Mach. ， 2017 ， 48 （ S1 ）： 422 - 428 . （in Chinese）

HUANG M ， WANG Q G ， ZHANG M ， et al . Prediction of color and moisture content for vegetable soybean during drying using hyperspectral imaging technology ［J］. J. Food Eng. ， 2014 ， 128 ： 24 - 30 . doi: 10.1016/j.jfoodeng.2013.12.008 http://dx.doi.org/10.1016/j.jfoodeng.2013.12.008

FENG L ， ZHU S S ， LIU F ， et al . Hyperspectral imaging for seed quality and safety inspection： a review ［J］. Plant Methods ， 2019 ， 15 （ 1 ）： 91 . doi: 10.1186/s13007-019-0476-y http://dx.doi.org/10.1186/s13007-019-0476-y

WANG L ， SUN D W ， PU H B ， et al . Application of hyperspectral imaging to discriminate the variety of maize seeds ［J］. Food Anal. Methods ， 2016 ， 9 （ 1 ）： 225 - 234 . doi: 10.1007/s12161-015-0160-4 http://dx.doi.org/10.1007/s12161-015-0160-4

ZHANG J ， DAI L M ， CHENG F . Corn seed variety classification based on hyperspectral reflectance imaging and deep convolutional neural network ［J］. J. Food Meas. Charact. ， 2021 ， 15 （ 1 ）： 484 - 494 . doi: 10.1007/s11694-020-00646-3 http://dx.doi.org/10.1007/s11694-020-00646-3

FANG X Y ， HU X N ， LI B L ， et al . Orbital angular momentum-mediated machine learning for high-accuracy mode-feature encoding ［J］. Light： Sci. Appl. ， 2024 ， 13 （ 1 ）： 49 . doi: 10.1038/s41377-024-01386-5 http://dx.doi.org/10.1038/s41377-024-01386-5

WANG Y F ， LIAO K ， ZHANG K ， et al . Reconfigurable versatile integrated photonic computing chip ［J］. eLight ， 2025 ， 5 （ 1 ）： 20 . doi: 10.1186/s43593-025-00098-6 http://dx.doi.org/10.1186/s43593-025-00098-6

XU P ， YANG R B ， ZENG T W ， et al . Varietal classification of maize seeds using computer vision and machine learning techniques ［J］. J. Food Process Eng. ， 2021 ， 44 （ 11 ）： e13846 . doi: 10.1111/jfpe.13846 http://dx.doi.org/10.1111/jfpe.13846

王驰，沈晨，黄庆，等 . 夜间动物图像自监督学习增强与检测方法［J］. 中国光学（中英文）， 2024 ， 17 （ 5 ）： 1087 - 1097 . doi: 10.37188/co.2024-0011 http://dx.doi.org/10.37188/co.2024-0011

WANG C ， SHEN C ， HUANG Q ， et al . Self-supervised learning enhancement and detection methods for nocturnal animal images ［J］. Chin. Opt. ， 2024 ， 17 （ 5 ）： 1087 - 1097 . （in Chinese） . doi: 10.37188/co.2024-0011 http://dx.doi.org/10.37188/co.2024-0011

JAVANMARDI S ， MIRAEI ASHTIANI S H ， VERBEEK F J ， et al . Computer-vision classification of corn seed varieties using deep convolutional neural network ［J］. J. Stored Prod. Res. ， 2021 ， 92 ： 101800 . doi: 10.1016/j.jspr.2021.101800 http://dx.doi.org/10.1016/j.jspr.2021.101800

LI J ， XU F ， SONG S Z ， et al . A maize seed variety identification method based on improving deep residual convolutional network ［J］. Front. Plant Sci. ， 2024 ， 15 ： 1382715 . doi: 10.3389/fpls.2024.1382715 http://dx.doi.org/10.3389/fpls.2024.1382715

BI C G ， HU N ， ZOU Y Q ， et al . Development of deep learning methodology for maize seed variety recognition based on improved swin transformer ［J］. Agronomy ， 2022 ， 12 （ 8 ）： 1843 . doi: 10.3390/agronomy12081843 http://dx.doi.org/10.3390/agronomy12081843

Views

163

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

CORM-YOLO： Pedestrian detection algorithm in low-light conditions for traffic environments

Lightweight detection of foreign objects on road surfaces in challenging driving environments

Retinal micro-vessel segmentation based on wavelet and coarse probability prior dual-branch network

PCB defect detection based on ZYNQ and YOLO26n

Quantitative relationship between flexoelectric coefficient difference and flicker in low-power fringe-field switching liquid crystal displays

Related Author

GENG Lijie

HOU Lingqi

FU Zhenxiang

BI Chenglong

ZHU Huajie

LIU Xuanyi

HOU Xingsong

YANG Kun

Related Institution

School of Electronic and Information， Henan Province Optoelectronic Sensing and Information Processing Engineering Technology Research Center，Zhengzhou University of Light Industry

College of Mechanical and Electronic Engineering， Northwest A&F University， Yangling

Henan Jiushenghe Rinke Seed Industry Co.， Ltd.

School of Mechanical and Electronic Engineering， Wuhan University of Technology

Hubei Tianwei Automobile Co. Ltd.

AI问答

Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China 130033 Postal code：130033
Tel：0431-86176059 Email：yjxs@ciomp.ac.cn
Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17 京公网安备11010802024621
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰