基于YOLOv11-FS模型的柑橘花粉活力检测

刘力源; 张学林; 陈洪; 李伟夫; 廖健华; 解凯东; 伍小萌; 郭文武; 陈耀晖

doi:10.13300/j.cnki.hnlkxb.2026.03.007

您当前的位置：

首页 >

文章列表页 >

基于YOLOv11-FS模型的柑橘花粉活力检测

农业人工智能专题 | 更新时间：2026-11-05

- 基于YOLOv11-FS模型的柑橘花粉活力检测
- YOLOv11-FS model-based detection of pollen viability in citrus
- 华中农业大学学报 2026年45卷第3期页码：77-86
- 作者机构：
  
  1.华中农业大学信息学院，武汉 430070
  2.果蔬园艺作物种质创新与利用全国重点实验室，武汉 430070
  3.华中农业大学工学院，武汉 430070
- 作者简介：
  
  刘力源，E-mail：llys@webmail.hzau.edu.cn
  陈耀晖，E-mail：yaohui.chen@mail.hzau.edu.cn
- 基金信息：
  
  国家重点研发计划项目(2023YFD1200103);教育部学科突破先导项目(JYB2025XDXM701);中央高校基本业务费专项(2662024SZ002)
- DOI：10.13300/j.cnki.hnlkxb.2026.03.007
  中图分类号： TP391.4;S666
- 收稿：2025-07-02，
  
  纸质出版：2026-05-25
- 稿件说明：
移动端阅览
刘力源，张学林，陈洪，等.基于YOLOv11-FS模型的柑橘花粉活力检测［J］.华中农业大学学报，2026，45（3）：77‐86.

LIU Liyuan,ZHANG Xuelin,CHEN Hong,et al.YOLOv11-FS model-based detection of pollen viability in citrus[J].Journal of Huazhong Agricultural University,2026,45(03):77-86.
刘力源，张学林，陈洪，等.基于YOLOv11-FS模型的柑橘花粉活力检测［J］.华中农业大学学报，2026，45（3）：77‐86. DOI： 10.13300/j.cnki.hnlkxb.2026.03.007.

LIU Liyuan,ZHANG Xuelin,CHEN Hong,et al.YOLOv11-FS model-based detection of pollen viability in citrus[J].Journal of Huazhong Agricultural University,2026,45(03):77-86. DOI： 10.13300/j.cnki.hnlkxb.2026.03.007.

摘要

为促进无核柑橘品种培育、满足高品质柑橘生产需求，提出一种新的花粉活力检测模型YOLOv11-FS。通过人工采集和专家标注的方式，构建了适用于花粉活力检测的数据集，改进了YOLOv11深度神经网络；检测过程中，采用Focal-EIOU损失函数替换YOLOv11的EIOU损失，改善对不均衡样本的检测性能，结合Soft-NMS提升检测框精度，克服柑橘花粉颗粒抱团明显、体积较小且背景复杂等挑战。结果显示，改进的YOLOv11-FS模型在花粉检测任务方面表现优秀，预测的花粉活力与真实的花粉活力误差仅为0.70百分点，可育花粉和不可育花粉检测的召回率、精确率和

-score分别达98.76%、99.67%、99.22%和94.87%、98.89%、96.84%，满足花粉活力检测基本需求。以上结果表明，该检测方法可实现柑橘花粉活力的高效、精准识别。

Abstract

A dataset of detecting pollen viability in citrus was constructed to promote the breeding of citrus seedless varieties and meet the demand for high-quality production of citrus. A new model of detecting pollen vitality was proposed by improving the YOLOv11 deep neural network. A dataset suitable for detecting pollen vitality was established through manual collection and expert annotation. The YOLOv11 deep neural network was improved to develop the YOLOv11-FS model. The Focal‑EIOU loss function was used to replace the EIOU loss of YOLOv11 to enhance detection performance on imbalanced samples during the process of detection. Soft‑NMS was incorporated to impr

ove the precision of detection boxes， overcoming challenges including obvious clustering， small size， and complex backgrounds of pollen grains in citrus. Results showed that the improved YOLOv11‑FS model performed excellently in the task of detecting pollen， the predicted pollen viability deviates from the true pollen viability by only 0.70 percentage points. The recall， precision， and

⁃score of detecting fertile pollen and sterile pollen reached 98.76%， 99.67%， 99.22%， and 94.87%， 98.89%， 96.84%， respectively， meeting the basic needs of detecting pollen vitality. It is indicated that the method of detection establshed can achieve efficient and accurate identification of pollen vitality in citrus.

关键词

Keywords

references

邓秀新，彭抒昂．柑橘学［M］.北京：中国农业出版社，2013. DENG X X，PENG S A.Citrology［M］.Beijing：China Agriculture Press，2013 （in Chinese） .

郭文武，叶俊丽，邓秀新.新中国果树科学研究70年：柑橘［J］.果树学报，2019，36（10）：1264-1272. GUO W W，YE J L，DENG X X.Fruit scientific research in New China in the past 70 years： Citrus ［J］.Journal of fruit science，2019，36（10）：1264-1272 （in Chinese with English abstract） .

LI S Q，YANG D C，ZHU Y G.Characterization and use of male sterility in hybrid rice breeding［J］.Journal of integrative plant biology，2007，49（6）：791-804.

VARDI A，LEVIN I，CARMI N.Induction of seedlessness in citrus：from classical techniques to emerging biotechnological approaches［J］.Journal of the American Society For Horticultural Science，2008，133（1）：117-126.

ZHENG B B，FANG Y N，PAN Z Y，et al.iTRAQ-based quantitative proteomics analysis revealed alterations of carbohydrate metabolism pathways and mitochondrial proteins in a male sterile cybrid pummelo［J］.Journal of proteome research，2014，13（6）：2998-3015.

LIANG M，CAO Z H，ZHU A D，et al.Evolution of self-compatibility by a mutant Sm - RNase in citrus［J］.Nature plants，20 20，6（2）：131-142.

SCOTT R J ， SPIELMAN M， DICKINSON H G. Stamen structure and function［J］.The plant cell，2004，16（suppl）：S46-S60.

WANG R，SHI Y C，XIE K D，et al.The citrus somatic cybrids exhibit various degrees of male sterility conferred by mitochondrial genome of Satsuma mandarin （ Citrus unshiu ）［J/OL］.Scientia horticulturae，2024，338：113624［2025-07-02］.https：//doi.org/10.1016/j.scienta.2024.113624.

ESCRICHE I，JUAN-BORRÁS M，VISQUERT M，et al.An overview of the challenges when analysing pollen for monofloral honey classification［J/OL］.Food control，2023，143：109305［2025-07-02］.https：//doi.org/10.1016/j.foodcont.2022.109305.

TRAVIESO C M， BRICENO J C， TICAY-RIVAS J R， et al. Pollen classification based on contour features ［C］//Proceedings of the 15th IEEE International Conference on Intelligent Engineering Systems. June 23-25，2011 ，Poprad， Slovakia. Poprad：IEEE，2011：17-21.

RODRIGUEZ-DAMIAN M，CERNADAS E，FORMELLA A，et al.Pollen classification using brightness-based and shape-based descriptors［C］//Proceedings of the 17th International Conference on Pattern Recognition. August 26-26，2004 ，Cambridge，UK. Cambridge：IEEE，2004：212-215.

SCHARRING S， RONNEBERGER O， KOCH W. Automatic pollen recognition-developments and perspectives automatische pollenerkennung-entwicklungen und perspektiven ［J］. Nachrichtenblatt des deutschen pflanzenschutzdienstes， 2006，58（11）：309-314.

PEDERSEN B，BAILEY D G，HODGSON R M，et al.Model and feature selection for the classification of dark field pollen images using the classifynder system［C］//2017 International Conference on Image and Vision Computing New Zealand （IVCNZ）. December 4-6，2017 ，Christchurch，New Zealand.Christchurch：IEEE，2017：1-5.

DEL POZO-BAÑOS M，TICAY-RIVAS J R，CABRERA-FALCN J，et al.Image processing for pollen classification［M/OL］//LAMEED G A. Biodiversity enrichment in a diverse world.［S.l.］：InTechOpen，2012.（2012-08-29）［2025-07-02］.https：//www.intechopen.com/chapters/38677.

DAOOD A，RIBEIRO E，BUSH M.Pollen grain recognition using deep learning［C］//Advances in Visual Computing. ISVC 2016. December 12-14， 2016 ，Las Vegas， NV， USA.Cham：Springer， 2016：321-330.

谢永华，徐赵飞，范文晓.基于高斯尺度空间粗糙度描述子的花粉图像分类识别［J］.计算机应用，2015，35（7）：2039-2042. XIE Y H，XU Z F，FAN W X.Pollen image classification and recognition based on Gaussian scale-space roughness descriptor［J］.Journal of computer applications，2015，35（7）：2039-2042 （in Chinese with English abstract） .

朱延刚.基于多特征融合的花粉图像识别算法研究［D］.南京：南京信息工程大学，2018.ZHU Y G.Research on pollen image recognition al gorithm based on multi-feature fusion［D］.Nanjing：Nanjing University of Information Science & Technology，2018 （in Chinese with English abstract） .

MAHMOOD T，CHOI J，PARK K R.Artificial intelligence-based classification of pollen grains using attention-guided pollen features aggregation network［J］.Journal of King Saud University computer and information sciences，2023，35（2）：740-756.

VALIENTE J M，JUAN-BORRÁS M，LÓPEZ-GARCÍA F，et al.Automatic pollen recognition using convolutional neural networks：the case of the main pollens present in Spanish citrus and rosemary honey［J/OL］.Journal of food composition and analysis，2023，123：105605［2025-07-02］.https：//doi.org/10.1016/j.jfca.2023.105605.

VALIENTE J M，MARTÍN-OSUNA J J，PERAL A M，et al.Assisting monofloral honey classification by automated pollen identification based on convolutional neural networks［J/OL］.Ecological informatics，2025，90：103340［2025-07-02］.https：//doi.org/10.1016/j.ecoinf.2025.103340.

杨倩，陈欣，张平冬，等.杨树2n花粉形态判别方法的比较［J］.东北林业大学学报，2015，43（2）：33-35.YANG Q，CHEN X，ZHANG P D，et al.Discrimination methods in 2n pollen in Populus ［J］.Journal of Northeast Forestry University，2015，43（2）：33-35 （in Chinese with English abstract） .

KHANAM R， HUSSAIN M. YOLOv11： An overview of the key architectural enhancements［PP/OL］. arXiv，2024：2410.17725［2025-07-02］.https：//doi.org/10.48550/arXiv.2410.17725.

ZHANG Y F，REN W Q，ZHANG Z，et al.Focal and efficient IOU loss for accurate bounding box regression［J］. Neurocomputing，2022，506：146-157.

BODLA N，SINGH B，CHELLAPPA R，et al.Soft-NMS：improving object detection with one line of code［PP/OL］.arXiv，2017：1704.04503［2025-07-02］. https：//doi.org/10.48550/arXiv.1704.04503.

TIAN Y N，YANG G D，WANG Z，et al.Apple detection during different growth stages in orchards using the improved YOLO-V3 model［J］.Computers and electronics in agriculture，2019，157：417-426.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

柑橘不同接穗对枳砧根际土壤理化性质与微生物群落结构的影响

胺鲜酯对椪柑果实品质及糖酸积累相关基因表达的影响

负载量对大雅柑果实产量、品质和翌年结果的影响

盛花期结果枝短截疏果对‘爱媛28号’杂柑果实品质及糖酸代谢基因表达的影响

大豆灰斑病病原变异、抗性挖掘与智能防控