1.江西农业大学软件学院,南昌330045
2.江西农业大学计算机与信息工程学院,南昌330045
黄带娣,E-mail:2871774998@qq.com
杨红云,E-mail:nc_yhy@163.com
收稿:2025-06-25,
纸质出版:2026-05-25
移动端阅览
黄带娣,杨红云,孙爱珍,等.基于MobileNetV3编码与U-Net多尺度解码融合的水稻磷素营养诊断[J].华中农业大学学报,2026,45(3):34‐44.
HUANG Daidi,YANG Hongyun,SUN Aizhen,et al.A rice phosphorus nutrition diagnosis method based on the fusion of MobileNetV3 encoding and U-Net multi-scale decoding[J].Journal of Huazhong Agricultural University,2026,45(03):34-44.
黄带娣,杨红云,孙爱珍,等.基于MobileNetV3编码与U-Net多尺度解码融合的水稻磷素营养诊断[J].华中农业大学学报,2026,45(3):34‐44. DOI: 10.13300/j.cnki.hnlkxb.2026.03.003.
HUANG Daidi,YANG Hongyun,SUN Aizhen,et al.A rice phosphorus nutrition diagnosis method based on the fusion of MobileNetV3 encoding and U-Net multi-scale decoding[J].Journal of Huazhong Agricultural University,2026,45(03):34-44. DOI: 10.13300/j.cnki.hnlkxb.2026.03.003.
为实现水稻叶片磷素营养状态的高效、精准诊断,提出一种融合轻量化编码器与改进U-Net解码结构的图像分类模型MobileNetV3_U-Net。以MobileNetV3_large作为轻量化编码器,提取多尺度深层语义特征,并加载预训练权重以加速收敛;解码器采用非对称多尺度上采样与跳跃连接逐级融合浅层特征,实现局部细节与全局语义的高效整合。与经典对称U-Net不同,MobileNetV3_U-Net模型通过逐级调整解码器的通道数和空间分辨率,更紧凑且专注于增强叶片细粒度特征表征,从而提升对磷素缺失特征的敏感性。最终通过自适应平均池化与全连接分类器完成水稻磷素营养状态判别。在水稻分蘖期和拔节期测试集,MobileNetV3_U-Net的准确率分别达到93.33%和87.40%,整体性能优于MobileNetV3_large、GhostNet、ShuffleNet和EfficientNet_b1等轻量化模型。在Plant Village公共数据集的葡萄叶片实验中,模型仍保持较高准确率,验证了其跨场景泛化能力。结果表明,MobileNetV3_U-Net在保持轻量化与高效部署的同时,通过改进U-Net解码结构实现多尺度局部与全局特征的深度融合,可为水稻磷素营养智能诊断提供一种高效、稳健且可推广的技术方案。
Accurate assessment of phosphorus nutrition is essential for optimizing rice growth and improving fertilizer management.This study proposes a lightweight image classification model, MobileNetV3_U-Net, that integrates a MobileNetV3_large encoder with an improved U-Net decoder for efficient diagnosis of rice leaf phosphorus status.The encoder leverages pretrained weights to accelerate convergence and extracts multi-scale semantic representations, while the decoder adopts an asymmetric multi-scale upsampling strategy with skip connections to progressively fuse shallow and deep features.Unlike the conventional symmetric U-Net, the redesigned decoder adjusts channel width and spatial resolution in a compact manner, enhancing fine-grained feature characterization and sensitivity to phosphorus deficiency symptoms.The classification is completed by adaptive average pooling and a fully connected layer.On test sets from the rice tillering and jointing stages, MobileNetV3_U-Net achieved accuracies of 93.33% and 87.40%, respectively, surpassing representative lightweight models including MobileNetV3_large, GhostNet, ShuffleNetV2, and EfficientNet_b1.Furthermore, experiments on the Plant Village dataset using grape leaf images confirmed its cross-domain generalization capability.MobileNetV3_U-Net demonstrates a favorable balance between lightweight design and diagnostic accuracy.By enhancing multi-scale feature fusion through a tailored decoder structure, it provides an efficient, robust, and transferable framework for intelligent diagnosis of rice phosphorus nutrition.
王伯仁,徐明岗,文石林,等.长期施肥对红壤旱地磷组分及磷有效性的影响[J].湖南农业大学学报(自然科学版),2002,28(4):293-297.WANG B R,XU M G,WEN S L,et al.The effects of long term fertilization on chemical fractions and availability of inorgauic phosphate in red soil upland[J].Journal of Hunan Agricultural University(natural sciences edition),2002,28(4):293-297 (in Chinese with English abstract) .
KHAN F,SIDDIQUE A B,SHABALA S,et al.Phosphorus plays key roles in regulating plants’ physiological responses to abiotic stresses[J/OL].Plants,2023,12(15):2861[2025-06-25].https://doi.org/10.3390/plants12152861.
ZHANG D K,LI Y Y,LI H R,et al.Phosphorus application rates affect the grain yields of different phosphorus-tolerant rice cultivars by regulating grain filling and leaf senescence characteristics[J].Plant,soil and environment,2025,71(5):363-380.
JULIA C,WISSUWA M,KRETZSCHMAR T,et al.Phosphorus uptake,partitioning and redistribution during grain filling in rice[J].Annals of botany,2016,118(6):1151-1162.
马晓诺,侯星宇,蔡红梅.长江中下游主栽水稻磷高效品种筛选及其生理生化特征[J].华中农业大学学报,2025,44(5):13‐24. MA X N ,HOU X Y,CAI H M,et al.Screening phosphorus-efficient rice varieties mainly planted in middle and lower reaches of Yangtze River and its physiological and biochemical characteristics[J].Journal of Huazhong Agricultural University,2025,44(5):13‐24 (in Chinese with English abstract) .
吕真真,刘秀梅,侯红乾,等.长期不同施肥对红壤性水稻土磷素及水稻磷营养的影响[J]. 植物营养与肥料学报,2019,25(8):1316-1324. LÜ Z Z,LIU X M,HOU H Q,et al.Effects of long-term fertilizations on soil phosphorus and its supply to rice in red paddy soil[J].Journal of plant nutrition and fertilizers,2019,25(8):1316-1324 (in Chinese with English abstract) .
ROSE T,KRETZSCHMAR T,LIU L,et al.Phosphorus deficiency alters nutrient accumulation patterns and grain nutritional quality in rice[J/OL].Agronomy,2016,6(4):52[2025-06-25].https://doi.org/10.3390/agronomy6040052.
王飞,李清华,林诚,等.南方低产黄泥田与高产灰泥田基础地力的差异[J].植物营养与肥料学报,2019,25(5):773-781. WANG F,LI Q H,LIN C,et al.Differences of basic soil productivity between yellow-mud and grey-mud paddy fields in Southern China[J].Journal of plant nutrition and fertilizers,2019,25(5):773-781 (in Chinese with English abstract) .
杨红云,周琼,杨珺,等.水稻氮素营养诊断方法研究进展[J].中国稻米,2020,26(2):5-8.YANG H Y ,ZHOU Q,YANG J,et al.Research progress on nitrogen nutrition diagnosis methods of rice[J].China rice,2020,26(2):5-8 (in Chinese with English abstract) .
侯云鹏,韩立国,孔丽丽,等.不同施氮水平下水稻的养分吸收、转运及土壤氮素平衡[J].植物营养与肥料学报,2015,21(4):836-845.HOU Y P,HAN L G,KONG L L,et al.Nutrient absorption,translocation in rice and soil nitrogen equilibrium under different nitrogen application doses[J].Journal of plant nutrition and fertilizers,2015,21(4):836-845 (in Chinese with English abstract) .
熊俊涛,戴森鑫,区炯洪,等.基于深度学习的大豆生长期叶片缺素症状检测方法[J].农业机械学报,2020, 51(1):195-202.XIONG J T,DAI S X,OU J H,et al.Leaf deficiency symptoms detection method of soybean based on deep learning[J].Transactions of the CSAM,2020,51(1):195-202 (in Chinese with English abstract) .
杨红云,郭紫微,郭高飞,等.基于Stacking集成卷积神经网络的水稻氮素营养诊断[J].植物营养与肥料学报,2023,29(3):573-581.YANG H Y,GUO Z W,GUO G F,et al.Rice nitrogen nutrition diagnosis based on Stacking integrated convolutional neural network[J].Plant nutrition and fertilizer science,2023,29(3):573-581 (in Chinese with English abstract) .
钱政,杨孙哲,张国卿,等.基于卷积神经网络的水稻氮素营养诊断[J].中国农业科技导报(中英文),2023,25(9):113-121.QIAN Z,YANG S Z,ZHANG G Q,et al.Rice nitrogen nutrition diagnosis based on convolutional neural network[J].Journal of agricultural science and technology,2023,25(9):113-121 (in Chinese with English abstract) .
黄淑梅,杨红云,孔杰,等.基于改进ShuffleNet V2的水稻磷素营养诊断方法[J].中国稻米,2025,31(2):20-28.HUANG S M,YANG H Y,KONG J,et al.Rice phosphorus nutrition diagnosis method based on improved ShuffleNet V2[J].China rice,2025,31(2):20-28 (in Chinese with English abstract) .
BISHWAS R,YASMIN S,TURZO N A.Relative comparison of K -means clustering segmented rice leaves image based nitrogen,phosphorus,and potassium nutrient deficiency classification using convolutional neural network[C]//2021 International Conference on Science & Contemporary Technologies (ICSCT).August 5-7,2021.Dhaka:IEEE,2021:1-6.
张林朋,杨红云,钱政,等.基于改进的VGG16网络和迁移学习的水稻氮素营养诊断[J].中国农业大学学报,2023,28(6):219-229.ZHANG L P,YANG H Y,QIAN Z,et al.Nitrogen nutrition diagnosis in rice based on improved VGG16 network and transfer learning[J].Journal of China Agricultural University,2023,28(6):219-229 (in Chinese with English abstract) .
吴正,杨红云,孙爱珍,等.基于PND-Net模型的水稻钾素胁迫程度诊断识别[J].江西农业大学学报,2024,46(6):1654-1667.WU Z,YANG H Y,SUN A Z,et al.Diagnosis and recognition of potassium stress degree in rice based on PND-Net model[J].Acta Agriculturae Universitatis Jiangxiensis (natural sciences edition),2024,46(6):1654-1667 (in Chinese with English abstract) .
孔杰,杨红云,黄淑梅,等.基于残差神经网络的水稻氮磷钾元素营养诊断[J].中国农业大学学报,2025,30(2):163-175.KONG J,YANG H Y,HUANG S M,et al.Nutritional diagnosis of nitrogen,phosphorus and potassium in rice based on the residual neural network[J].Journal of China Agricultural University,2025,30(2):163-175 (in Chinese with English abstract) .
SMITH B T,WANG L.Comparison of DenseNet201, ResNet50, VGG16, and xception models for soybean disease leaf image classification[J].Journal of agricultural engineering and technology, 45(3), 115-126.
YANG X, ZHANG L, LIU Y.Improving GoogLeNet with ECA and residual connections for image classification[J].Journal of visual communication and image representation, 2019,62: 234-245.
RONNEBERGER O,FISCHER P,BROX T.U-net:convolutional networks for biomedical image segmentation[M]//NAVAB N, HORNEGGER J, WELLS W.Medical image computing and computer-assisted intervention.Cham:Springer International Publishing,2015:234-241.
LIU W F,YU L,LUO J X.A hybrid attention-enhanced DenseNet neural network model based on improved U-Net for rice leaf disease identification[J/OL].Frontiers in plant science,2022,13:922809[2025-06-25].https://doi.org/10.3389/fpls.2022.922809.
WANG Y,ZHANG D F,DAI G M.Classification of high resolution satellite images using improved U-Net[J].International journal of applied mathematics and computer science,2020,30(3):399-413.
王军,张霁云,程勇.基于边缘特征和注意力机制的图像语义分割[J].计算机系统应用,2024,33(7):63-73.WANG J,ZHANG J Y,CHENG Y.Image semantic segmentation based on edge features and attention mechanism[J].Computer systems and applications,2024,33(7):63-73 (in Chinese with English abstract) .
WU F Z,WU J S,KONG Y Y,et al.Multiscale low-frequency memory network for improved feature extraction in convolutional neural networks[J].Proceedings of the AAAI conference on artificial intelligence,2024,38(6):5967-5975.
易振通,吴瑰,官端正,等.轻量化卷积神经网络的研究综述[J].工业控制计算机,2022,35(10):109-111.YI Z T,WU G,GUAN D Z,et al.Survey of research on lightweight convolutional neural networks[J].Industrial control computer,2022,35(10):109-111 (in Chinese with English abstract) .
WU B C,KEUTZER K,DAI X L,et al.FBNet:hardware-aware efficient ConvNet design via differentiable neural architecture search[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).June 15-20,2019.Long Beach:IEEE,2019:10726-10734.
SHORTEN C,KHOSHGOFTAAR T M.A survey on image data augmentation for deep learning[J/OL].Journal of big data,2019,6:60[2025-06-25].https://doi.org/10.1186/s40537-019-0197-0.
HOWARD A,SANDLER M,CHEN B,et al.Searching for MobileNetV3[C]//2019 IEEE/CVF International Conference on Computer Vision (ICCV),October 27-November 2,2019.Seoul:IEEE,2019:1314-1324.
0
浏览量
4
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621