[1]张园,胡峻峰,刘大洋.基于URMSM网络的玉米叶部病斑分割[J].江苏农业科学,2026,54(4):205-212.
 Zhang Yuan,et al.Study on maize leaf lesion segmentation based on URMSM network[J].Jiangsu Agricultural Sciences,2026,54(4):205-212.
点击复制

基于URMSM网络的玉米叶部病斑分割()

《江苏农业科学》[ISSN:1002-1302/CN:32-1214/S]

卷:
第54卷
期数:
2026年第4期
页码:
205-212
栏目:
病害智能检测
出版日期:
2026-02-20

文章信息/Info

Title:
Study on maize leaf lesion segmentation based on URMSM network
作者:
张园胡峻峰刘大洋
东北林业大学计算机科学与控制工程学院,黑龙江哈尔滨 150040
Author(s):
Zhang Yuanet al
关键词:
玉米叶部病害病斑分割复杂背景干扰语义分割UR-MSM网络
Keywords:
-
分类号:
S126;S435.131;TP391.41
DOI:
-
文献标志码:
A
摘要:
针对复杂背景干扰下玉米叶部病斑图像分割精度不高的问题,提出一种改进的分割网络UR-MSM,以提升其在阴影、遮挡及光照变化等复杂场景下的分割准确性、稳定性与鲁棒性。本研究采用融合公共数据集与自建数据的混合数据集CGSDD进行模型训练与验证。UR-MSM网络以U-Net为基础架构,对以下3个关键方面作出改进:首先,使用SE-ResNet模块替换编码器中传统的双层卷积操作,以增强特征表征能力并有效缓解梯度消失问题,同时提升模型对光照变化的适应能力;其次,引入矩形自校准(RCM)模块,实现编码器不同层级特征的有效融合,增强网络对病斑前景区域的关注,从而解决由阴影造成的局部特征丢失问题;最后,设计空间与通道协同注意力(SCSA)机制,通过空间和通道2个维度的双向协同优化,减小语义歧义,提升特征选择能力。试验结果表明,UR-MSM网络有效提升了在复杂干扰情况下玉米病斑的分割性能。在测试集上,该模型的MIoU、宏F1分数分别达到88.20%、9348%,相较于原始U-Net基线模型分别提升4.72、3.00百分点。在与DeepLab v3+、PSPNet等其他主流分割模型的对比试验中,本研究提出的模型在处理阴影遮挡场景方面展现出明显优势;通过系统对比不同数据增强策略下的性能表现,进一步验证了该模型具有良好的泛化性与稳定性。本研究提出的UR-MSM网络在实现模型轻量化的前提下,通过对网络结构的针对性优化,显著提高了分割玉米叶部病斑图像的综合性能。
Abstract:
-

参考文献/References:

[1]体院艾. 玉米病虫害防治与种植技术应用研究[J]. 工程管理,2023,4(6):64-66,68.
[2]苏杰. 设施农业中玉米种植智能病虫害监测与防治技术[J]. 农业工程技术,2024,44(29):60-61.
[3]蒋广轩. 玉米主要病害及防治对策[J]. 种子科技,2025,43(7):167-169,172.
[4]吴忠. 玉米病虫害发生特点及防治措施研究[J]. 热带农业工程,2022,46(3):66-68.
[5]王秋林,舒雪锋,贺春平. 玉米病虫害防治及合理使用农药的措施[J]. 农业灾害研究,2023,13(1):16-18.
[6]法杰赵. 高产玉米种植技术及病虫害强化防治措施[J]. 工程管理,2023,4(4):52-54.
[7]Si Y,Xu H,Zhu X,et al. SCSA:Exploring the synergistic effects between spatial and channel attention[J]. Neurocomputing,2025,634(14):129866.
[8]Ni Z,Chen X,Zhai Y,et al. Context-guided spatial feature reconstruction for efficient semantic segmentation[C]// European Conference on Computer Vision.Cham:Springer Nature Switzerland,2024:239-255.
[9]He K,Zhang X,Ren S,et al. Deep residual learning for image recognition[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas:IEEE,2016:770-778.
[10]Hu J,Shen L,Sun G. Squeeze-and-excitation networks[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Salt Lake City:IEEE,2018:7132-7141.
[11]Yurtkulu S C,?瘙塁ahin Y H,Unal G.Semantic segmentation with extended DeepLabv3 architecture[C]// 2019 27th Signal Processing and Communications Applications Conference. Sivas:IEEE,2019:1-4.
[12]Zhou J,Hao M,Zhang D,et al. Fusion PSPnet image segmentation based method for multi-focus image fusion[J]. IEEE Photonics Journal,2019,11(6):1-12.
[13]Oktay O,Schlemper J,Folgoc L L,et al. Attention u-net:Learning where to look for the pancreas[J]. arXiv preprint arXiv:1804.03999,2018.
[14]戚海洲,吴敬兵,郭荣秋.基于改进DeepLabv3+的草坪区域语义分割方法[J]. 自动化与仪表,2025,40(4):132-136,141.
[15]Chen G,Li L,Dai Y,et al. AAU-net:an adaptive attention U-net for breast lesions segmentation in ultrasound images[J]. IEEE Transactions on Medical Imaging,2022,42(5):1289-1300.
[16]乔豪杰,王晓云. 基于Lab和全局阈值分割的水果图像分割算法[J]. 装备机械,2024,188(2):12-16+64.
[17]孙晨,王昕,蒋国臻. 基于Levy-SOA自适应阈值分割和改进引导滤波的NSST图像增强[J]. 控制工程,2024,31(7):1297-1304.
[18]粟长权,郭本华,魏一帆,等. 融合Canny边缘检测的多输出损失肺炎CT图像分割算法[J]. 现代计算机,2024,30(1):1-8,17.
[19]Karabagˇ C,Miller N,Garcia-Finana M,et al. Texture segmentation:an objective comparison between five traditional algorithms and a deep-learning U-Net architecture[J]. Applied Sciences,2019,9(18):3900.
[20]Luo Z F,Yuan Y,Liu Y F,et al. Semantic segmentation of agricultural images:a survey[J]. Information Processing in Agriculture,2024,11(2):172-186
[21]Zhu Q,Weng N,Fan L,et al. Enhancing environmental monitoring through multispectral imaging:the WasteMS dataset for semantic segmentation of lakeside waste[C]// International Conference on Multimedia Modeling.Singapore:Springer Nature Singapore,2025:362-372.
[22]Tran D P,Nguyen Q A,Pham V T,et al. Trans2Unet:neural fusion for nuclei semantic segmentation[C]// 2022 11th International Conference on Control,Automation and Information Sciences. Hanoi:IEEE,2022:583-588.
[23]李凯雨,张慧,马浚诚,等. 基于语义分割和可见光谱图的作物叶部病斑分割方法[J]. 光谱学与光谱分析,2023,43(4):1248-1253.
[24]芦碧波,梁迪,杨洁,等. 基于改进ENet的复杂背景下山药叶片图像分割方法[J]. 智慧农业(中英文),2024,6(6):109-120.
[25]孙茜容,王小鹏. 一种改进 ResUNet 的遥感影像建筑物提取方法[J]. Laser & Optoelectronics Progress,2025,62(8):0828003-0828003-11.
[26]姜文文,夏英. 改进U-Net的多尺度特征融合遥感图像语义分割网络[J]. 计算机科学,2025,52(5):212-219.
[27]胡磊,胡蓉,赵全友,等. 基于MATLAB的柑橘病虫害图像分割算法的研究[J]. 大众科技,2020,22(11):9-11.
[28]沈艳艳,赵玉涛,陈庚申,等. 玉米典型叶部病害高光谱识别及其烈度分类[J]. 智慧农业(中英文),2024,6(2):28-39.
[29]王营瑛,郑铖,董伟,等. 一种基于改进卷积神经网络的玉米病害高效识别模型[J]. 安徽科技学院学报,2023,37(4):96-104.
[30]Lu Y,Chen Y,Zhao D,et al. Graph-FCN for image semantic segmentation[C]// International Symposium on Neural Networks.Cham:Springer International Publishing,2019:97-105.
[31]徐婷宜,朱家明,李祥健. 基于全卷积网络的肝脏CT语义分割[J]. 软件工程,2020,23(6):20-22.
[32]Kalpana P,Chanti Y,Pareek P K.SE-Resnet152 model:early corn leaf disease identification and classification using feature based transfer learning technique[C]// 2023 International Conference on Evolutionary Algorithms and Soft Computing Techniques. Bengaluru:IEEE,2023:1-6.DOI:10.1109/EASCT59475.2023.10392328.
[33]Zhen J,Wang X,Li Y,et al. Detection of strawberry bloom phenology based on YOLO-RCMC and flower opening scale[J]. Smart Agricultural Technology,2025,12:101275.
[34]乔世成,潘春宇,白明宇,等. 基于改进YOLO v10n的石榴病害检测[J]. 沈阳农业大学学报,2025,56(4):93-102.
[35]Ahmad A,Saraswat D,Gamal A E,et al. CD&S dataset:handheld imagery dataset acquired under field conditions for corn disease identification and severity estimation[EB/OL]. (2021-10-22) [2025-09-18]. https://arxiv.org/abs/2110.12084.

相似文献/References:

[1]孙建伟.水涝胁迫对玉米细胞保护酶同工酶的影响[J].江苏农业科学,2013,41(04):85.
[2]刘荣,张卫建,齐华,等.密植型玉米“中单909”高产群体结构特征[J].江苏农业科学,2013,41(05):56.
 Liu Rong,et al.Study on high yield population structure of close planting maize cultivar “Zhongdan 909”[J].Jiangsu Agricultural Sciences,2013,41(4):56.
[3]沈浜凯,肖龙云,冯乃杰,等.黄腐酸和AM真菌对玉米幼苗抗旱性的影响[J].江苏农业科学,2013,41(05):64.
 Shen Bangkai,et al.Effects of fulvic acid and AM fungi on drought resistance of maize seedlings[J].Jiangsu Agricultural Sciences,2013,41(4):64.
[4]张金然,缑艳霞,孙丽鹏.固氮螺菌157对玉米、向日葵的促生长作用[J].江苏农业科学,2014,42(12):116.
 Zhang Jinran,et al.Effects of Azospirillum 157 on growth of maize and sunflower[J].Jiangsu Agricultural Sciences,2014,42(4):116.
[5]白小军,吴燕,牛艳,等.玉米中乙草胺和莠去津残留量GC-MS/MS分析法的建立[J].江苏农业科学,2014,42(11):334.
 Bai Xiaojun,et al().Establishment of GC-MS/MS analysis method of acetochlor and atrazine residues in maize[J].Jiangsu Agricultural Sciences,2014,42(4):334.
[6]邹晓威,王娜,刘芬,等.玉米抗病相关基因在玉米与玉米丝黑穗病菌、玉米黑粉病菌互作过程中的表达差异分析[J].江苏农业科学,2014,42(11):150.
 Zou Xiaowei,et al(0).Different expression of resistance-related genes between Sporisorium reilianum and Ustilago maydis interact with corn[J].Jiangsu Agricultural Sciences,2014,42(4):150.
[7]杨洪兴,陈静,陈艳萍.江苏省玉米机械化生产的发展及育种对策思考[J].江苏农业科学,2014,42(11):116.
 Yang Hongxing,et al().Development and breeding strategy of mechanized production of maize in Jiangsu Province[J].Jiangsu Agricultural Sciences,2014,42(4):116.
[8]张丽妍,霍剑锋,孟繁盛,等.不同肥料、施肥水平及施用方法对玉米产量、性状及效益的影响[J].江苏农业科学,2014,42(11):119.
 Zhang Liyan,et al (9).Effects of different fertilizers,fertilizer levels and fertilizing methods on yield,characters and benefit of maize[J].Jiangsu Agricultural Sciences,2014,42(4):119.
[9]王雷,崔震海,张立军.玉米C4型PEPC全长基因的克隆与表达载体构建[J].江苏农业科学,2014,42(11):26.
 Wang Lei,et al().Cloning and expression vector construction of full-length C4 type PEPC gene in maize[J].Jiangsu Agricultural Sciences,2014,42(4):26.
[10]雷恩,赵光明,刘艳红.不同稀释浓度松土保水剂对玉米营养生长的影响[J].江苏农业科学,2013,41(06):77.
 Lei En,et al.Effect of different dilutions of super absorbent polymer on vegetative growth of maize[J].Jiangsu Agricultural Sciences,2013,41(4):77.
[11]刘丽娟,刘仲鹏.基于改进BP算法的玉米叶部病害图像识别研究[J].江苏农业科学,2013,41(11):139.
 Liu Lijuan,et al.Image recognition of maize leaf diseases based on improved BP algorithm[J].Jiangsu Agricultural Sciences,2013,41(4):139.

备注/Memo

备注/Memo:
收稿日期:2025-08-13
基金项目:国家自然科学基金(编号:32202147);中央高校基本科研专项资金(编号:2572019BF09)。
作者简介:张园(2001—),女,山东单县人,硕士研究生,主要研究方向为模式识别。E-mail:nefu_zyy@126.com。
通信作者:胡峻峰,博士,副教授,主要研究方向为模式识别。E-mail:nefuhjf@126.com。
更新日期/Last Update: 2026-02-20