[1]胡健威,马慧敏,宁孝梅,等.基于改进YOLOv8的无人机图像玉米幼苗检测[J].江苏农业学报,2025,(06):1179-1187.[doi:doi:10.3969/j.issn.1000-4440.2025.06.014]
　HU Jianwei,MA Huiming,NING Xiaomei,et al.Corn seedling detection in unmanned aerial vehicle images based on improved YOLOv8[J].,2025,(06):1179-1187.[doi:doi:10.3969/j.issn.1000-4440.2025.06.014]
点击复制

基于改进YOLOv8的无人机图像玉米幼苗检测()
分享到：

江苏农业学报[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2025年06期
页码:
1179-1187
栏目:
农业信息工程
出版日期:
2025-06-30

文章信息/Info

Title:
Corn seedling detection in unmanned aerial vehicle images based on improved YOLOv8
作者:
胡健威; 马慧敏; 宁孝梅; 代腾辉; 戴明宇; 王小申; 吴旖
(安徽农业大学信息与人工智能学院,安徽合肥230036)
Author(s):
HU Jianwei; MA Huiming; NING Xiaomei; DAI Tenghui; DAI Mingyu; WANG Xiaoshen; WU Yi
(School of Information and Artificial Intelligence, Anhui University of Agriculture, Hefei 230036, China)
关键词:
玉米幼苗; 无人机; YOLOv8; MultiSEAM; MetaNeXtStage; 损失函数
Keywords:
corn seedling; unmanned aerial vehicle; YOLOv8; MultiSEAM; MetaNeXtStage; loss function
分类号:
S513
DOI:
doi:10.3969/j.issn.1000-4440.2025.06.014
文献标志码:
A
摘要:
无人机技术凭借其高效、精准的优势,在农业领域发挥着重要作用,被广泛应用于农田监测、精准施肥、病虫害防治等环节。然而,在无人机采集的玉米幼苗图像中,田间杂草等干扰物与玉米幼苗颜色相近,易导致YOLOv8模型误检和漏检；同时玉米幼苗间相互遮挡,也会影响模型检测精度。针对这些问题,本研究提出了YOLOv8+MultiSEAM+MetaNeXtStag+WIoU模型（简称YOLOv8-MMW模型）。该模型基于YOLOv8模型架构,首先在颈部网络引入MultiSEAM注意力机制,有效提升了模型在复杂场景下的特征提取能力；其次引入了InceptionNeXt中的MetaNeXtStage模块；在此基础上,采用Wise-IoU损失函数以提升模型精度。在测试集上的试验结果表明,YOLOv8-MMW模型精度和交并比阈值为0.50时的平均精度均值（mAP50）分别达到98.9%和89.6%,较原始YOLOv8n模型分别提升了6.1个百分点和2.4个百分点。本研究提出的YOLOv8_MMW模型在复杂农田环境下表现出更强的鲁棒性,能够有效提升对无人机拍摄的玉米幼苗图像的检测准确率,为农业管理和监测提供了技术支持。
Abstract:
Unmanned aerial vehicle technology, with its high efficiency and precision, has been playing an important role in the agricultural field and has been widely applied in farmland monitoring, precision fertilization, and pest and disease control. However, in the corn seedling images collected by unmanned aerial vehicle, weeds and other interferents in the field have colors similar to corn seedlings, which can easily lead to false detection and missed detection in the YOLOv8 model. Additionally, the occlusion between corn seedlings can also affect the detection accuracy of the model. In response to these problems, this study proposed the YOLOv8+MultiSEAM+MetaNeXtStag+WIoU (YOLOv8-MMW) model. Based on the YOLOv8 model architecture, this model first introduced the MultiSEAM attention mechanism into the neck network, which effectively enhanced the model’s feature extraction ability in complex scenes. Secondly, it incorporated the MetaNeXtStage module from InceptionNeXt. On this basis, the Wise-IoU loss function was adopted to improve model accuracy. The experimental results on the test set showed that the accuracy and the mean average precision at an intersection-over-union threshold of 0.50 (mAP50) of the YOLOv8-MMW model reach 98.9% and 89.6%, respectively, which were 6.1 percentage points and 2.4 percentage points higher than those of the original YOLOv8n model. The YOLOv8-MMW model proposed in this study demonstrates stronger robustness in complex farmland environments and can effectively improve the detection accuracy of corn seedling images captured by unmanned aerial vehicle, and can provide technical support for agricultural management and monitoring.

参考文献/References:

[1]仇焕广,栾昊,李瑾,等. 风险规避对农户化肥过量施用行为的影响[J]. 中国农村经济,2014(3):85-96.
[2]徐云姬,顾道健,张博博,等. 玉米果穗不同部位籽粒激素含量及其与胚乳发育和籽粒灌浆的关系[J]. 作物学报,2013,39(8):1452-1461.
[3]赵久然,王荣焕. 中国玉米生产发展历程、存在问题及对策中国农业科技导报[J]. 中国农业科技导报,2013(3):1-6.
[4]陈源源,吕昌河,尚凯丽. 食物安全的内涵、指标与评价方法综述[J]. 中国农学通报,2017,33(22):158-164.
[5]WANG Y, XIE F, ZHAO C X, et al. Robust face recognition model based sample mining and loss functions[J]. Knowledge-Based Systems,2024,302:112330.
[6]ZHANG Y S, YAN S K, ZHANG L F, et al. Fast projected fuzzy clustering with anchor guidance for multimodal remote sensing imagery[J]. IEEE transactions on image processing,2024(33),4640-4653 .
[7]QIU C Q, TANG H, YANG Y C, et al. Machine vision-based autonomous road hazard avoidance system for self-driving vehicles[J]. Scientific Reports,2024,14(1):12178.
[8]JIANG H K, CHEN Q C, WANG R J, et al. SWFormer:a scale-wise hybrid CNN-Transformer network for multi-classes weed segmentation[J]. Journal of King Saud University-Computer and Information Sciences,2024,36(7):102144.
[9]MUQADDAS S, QURESHI W S, JABBAR H, et al. A comprehensive deep learning approach for harvest ready sugarcane pixel classification in Punjab,Pakistan using Sentinel-2 multispectral imagery[J]. Remote Sensing Applications:Society and Environment,2024,35:101225.
[10]周善良,李锐. 基于卷积神经网络的农作物病虫害识别研究综述[J]. 智慧农业导刊,2024,4(17):39-45.
[11]CHEN F, JI X Z, BAI M X, et al. Network analysis of different exogenous hormones on the regulation of deep sowing tolerance in maize seedlings[J]. Frontiers in Plant Science,2021,12:739101.
[12]MALIK M M, FAYYAZ A M, YASMIN M, et al. A novel deep CNN model with entropy coded sine cosine for corn disease classification[J]. Journal of King Saud University-Computer and Information Sciences,2024,36(7):102126.
[13]LIU J Q, HU Y X, SU Q F, et al. Semi-supervised one-stage object detection for maize leaf disease[J]. Agriculture,2024,14 (7):1140.
[14]李金瑞,杜建军,张宏鸣,等. 基于轻量化MLCE-RTMDet的人工去雄后玉米雄穗检测算法[J]. 农业机械学报,2024,55(11):1-14.
[15]PUYENBROECK E V, WOUTERS N, LEBLICQ T, et al. Detection of kernels in maize forage using hyperspectral imaging[J]. Computers and Electronics in Agriculture,2024,225:109336.
[16]MEI C L, XUE Y C, LI Q H, et al. Deep learning model based on molecular spectra to determine chlorpyrifos residues in corn oil[J]. Infrared Physics and Technology，2024,140:105402.
[17]FAN B K, LI Y, ZHANG R Y, et al. Review on the technological development and application of UAV systems[J]. Chinese Journal of Electronics,2020,29(2):5-13.
[18]ABEL B, PHILIPP L, RAMN I Y F, et al. Automatic UAV-based counting of seedlings in sugar-beet field and extension to maize and strawberry[J]. Computers and Electronics in Agriculture,2021,191:106493.
[19]CHEN J H, FU Y S, GUO Y H, et al. An improved deep learning approach for detection of maize tassels using UAV-based RGB images[J]. International Journal of Applied Earth Observation and Geoinformation,2024,130:103922.
[20]XIAO J, SUAB S A, CHEN X Y, et al. Enhancing assessment of corn growth performance using unmanned aerial vehicles (UAVs) and deep learning[J]. Measurement,2023, 214:112764.
[21]LV Z G, XU B Y, ZHONG L H, et al. Improved monitoring of southern corn rust using UAV-based multi-view imagery and an attention-based deep learning method[J]. Computers and Electronics in Agriculture,2024,224:109232.
[22]邬开俊,白晨帅,杜建军,等. 无人机影像的玉米植株中心检测模型和方法[J]. 计算机工程与应用,2024(6):1-17.
[23]NI X D, WANG F M, HUANG H, et al. A CNN-and self-attention-based maize growth stage recognition method and platform from UAV orthophoto images[J]. Remote Sensing,2024,16(14):2672.
[24]杨国亮,盛杨杨,洪鑫芳,等. 改进YOLOv8n的果园番茄目标检测算法[J]. 计算机工程与应用,2024,60(23):1-13.
[25]WANG Y D, ZHANG J, KAN M N, et al. Self-supervised equivariant attention mechanism for weakly supervised semantic segmentation[EB/OL].(2020-04-09)
[2024-10-30]. https://arxiv.org/abs/2004.04581.
[26]YU W H, ZHOU P, YAN S C, et al. InceptionNeXt:when inception meets ConvNeXt[EB/OL].(2023-03-29)
[2024-10-30]. https://arxiv.org/abs/2303.16900.
[27]TONG Z J, CHEN Y H, XU Z W, et al. Wise-IoU:bounding box regression loss with dynamic focusing mechanism[EB/OL].(2023-01-24)
[2024-10-30]. https://arxiv.org/abs/2301.10051.
[28]ZHANG Y F, REN W Q, ZHANG Z, et al. Focal and efficient IOU loss for accurate bounding box regression[EB/OL].(2021-01-20)
[2024-10-30]. https://arxiv.org/abs/2101.08158.context=cs.CV.
[29]ZHENG Z H, WANG P, LIU W, et al. Distance-IoU loss:faster and better learning for bounding box regression[EB/OL].(2019-11-19)
[2024-10-30]. https://arxiv.org/abs/1911.08287.
[30]CHAKROUN I, HABER T, ASHBY J T. SW-SGD:the sliding window stochastic gradient descent algorithm[J]. Procedia Computer Science,2017,108:2318-2322.

相似文献/References:

[1]于堃,单捷,王志明,等.无人机遥感技术在小尺度土地利用现状动态监测中的应用[J].江苏农业学报,2019,(04):853.[doi:doi:10.3969/j.issn.1000-4440.2019.04.015]
　YU Kun,SHAN Jie,WANG Zhi ming,et al.Land use status monitoring in small scale by unmanned aerial vehicles (UAVs) observations[J].,2019,(06):853.[doi:doi:10.3969/j.issn.1000-4440.2019.04.015]
[2]陶惠林,冯海宽,徐良骥,等.基于无人机高光谱遥感数据的冬小麦生物量估算[J].江苏农业学报,2020,(05):1154.[doi:doi:10.3969/j.issn.1000-4440.2020.05.012]
　TAO Hui-lin,FENG Hai-kuan,XU Liang-ji,et al.Winter wheat biomass estimation based on hyperspectral remote sensing data of unmanned aerial vehicle(UAV)[J].,2020,(06):1154.[doi:doi:10.3969/j.issn.1000-4440.2020.05.012]
[3]张先洁,孙国祥,汪小旵,等.基于超像素特征向量的果树冠层分割方法[J].江苏农业学报,2021,(03):724.[doi:doi:10.3969/j.issn.1000-4440.2021.03.023]
　ZHANG Xian-jie,SUN Guo-xiang,WANG Xiao-chan,et al.Segmentation method of fruit tree canopy based on super pixel feature vector[J].,2021,(06):724.[doi:doi:10.3969/j.issn.1000-4440.2021.03.023]
[4]郭松,常庆瑞,郑智康,等.基于无人机高光谱影像的玉米叶绿素含量估测[J].江苏农业学报,2022,38(04):976.[doi:doi:10.3969/j.issn.1000-4440.2022.04.014]
　GUO Song,CHANG Qing-rui,ZHENG Zhi-kang,et al.Estimation of maize chlorophyll content based on unmanned aerial vehicle (UAV) hyperspectral images[J].,2022,38(06):976.[doi:doi:10.3969/j.issn.1000-4440.2022.04.014]
[5]龚志远,李雪梅,李秋萍,等.兰州植物园植被春季物候无人机监测[J].江苏农业学报,2023,(08):1707.[doi:doi:10.3969/j.issn.1000-4440.2023.08.010]
　GONG Zhi-yuan,LI Xue-mei,LI Qiu-ping,et al.Unmanned aerial vehicle (UAV) monitoring of spring vegetation phenology in Lanzhou Botanical Garden[J].,2023,(06):1707.[doi:doi:10.3969/j.issn.1000-4440.2023.08.010]
[6]李瑞鑫,张宝林,潘丽杰,等.不同无人机飞行高度下玉米叶片叶绿素相对含量的无人机遥感反演及其指示叶位的识别[J].江苏农业学报,2024,(07):1234.[doi:doi:10.3969/j.issn.1000-4440.2024.07.010]
　LI Ruixin,ZHANG Baolin,PAN Lijie,et al.Unmanned aerial vehicle remote sensing inversion of relative chlorophyll content of maize leaves and identification of their indicator leaf at different flight altitudes[J].,2024,(06):1234.[doi:doi:10.3969/j.issn.1000-4440.2024.07.010]
[7]杨玉青,朱德泉,刘凯旋,等.基于改进的LSN-YOLOv8模型和无人机遥感图像的水稻稻曲病检测方法[J].江苏农业学报,2025,(05):905.[doi:doi:10.3969/j.issn.1000-4440.2025.05.009]
　YANG Yuqing,ZHU Dequan,LIU Kaixuan,et al.A method for rice false smut detection based on improved LSN-YOLOv8 model and unmanned aerial vehicle remote sensing images[J].,2025,(06):905.[doi:doi:10.3969/j.issn.1000-4440.2025.05.009]
[8]刘新侠,杨鑫宇,付春晓,等.基于无人机多光谱估算冬小麦作物系数[J].江苏农业学报,2025,(06):1169.[doi:doi:10.3969/j.issn.1000-4440.2025.06.013]
　LIU Xinxia,YANG Xinyu,FU Chunxiao,et al.Crop coefficient estimation of winter wheat based on unmanned aerial vehicle (UAV) multispectral data[J].,2025,(06):1169.[doi:doi:10.3969/j.issn.1000-4440.2025.06.013]

备注/Memo

备注/Memo:
收稿日期：2024-11-09基金项目：安徽省高校优秀青年人才支持项目(gxyq2022004)；安徽省研究生创新创业项目(2022cxcysj065)作者简介：胡健威(1998-),男,安徽合肥人,硕士研究生,研究方向为作物信息智能感知。(E-mail)jwhu@stu.ahau.edu.cn通讯作者：马慧敏,(E-mail)huiminma@ahau.edu.cn
更新日期/Last Update: 2025-07-16