Welcome to Acta Agronomica Sinica,

Acta Agronomica Sinica ›› 2026, Vol. 52 ›› Issue (5): 1401-1417.doi: 10.3724/SP.J.1006.2026.51089

• CROP GENETICS & BREEDING·GERMPLASM RESOURCES·MOLECULAR GENETICS • Previous Articles     Next Articles

Development of a deep learning-based image recognition system for major wheat diseases

He Wan-Long(), Geng Hong-Wei, Zhang Fei-Fei, Mikereayi·Ababaikere , Luo Zi-Yang, Li Peng-Cheng, Zhou Zhao-Yu, Cheng Yu-Kun*()   

  1. College of Agronomy, Xinjiang Agricultural University / Engineering Research Center for High-quality Special Wheat Crops, Xinjiang Agricultural University / Innovation Team for Xinjiang Wheat Industry System, Urumqi 830052, Xinjiang, China
  • Received:2025-10-28 Accepted:2026-01-22 Online:2026-05-12 Published:2026-02-04
  • Contact: *Cheng Yu-Kun, E-mail: ykcheng@126.com
  • Supported by:
    Major Science and Technology Project of the Xinjiang Uygur Autonomous Region(2024A02003-2)

Abstract:

Traditional methods for wheat disease identification rely heavily on manual surveys, which are labor-intensive and time-consuming. To address these limitations, this study proposes a recognition model, CBAM-MobileNet-V2, based on an enhanced MobileNet-V2 architecture. Designed for deployment on Android mobile platforms, the model provides agricultural practitioners with a fast, accurate, and user-friendly tool for in-field wheat disease diagnosis. The model offers several key advantages. First, it employs transfer learning to facilitate efficient feature extraction. Second, residual connections are integrated into the core feature extraction modules to enhance learning capacity, while average pooling layers within the inverted residual blocks are used for effective downsampling. Additionally, the CBAM attention mechanism is incorporated to further improve feature extraction and overall model performance. The optimal performance is achieved with a regularization coefficient of 0.005 and a learning rate of 0.010. The model is trained using the AdamW optimizer, which enhanced convergence and generalization. CBAM-MobileNet-V2 has a compact size of only 9.69 Mb and achieves an accuracy of 97.95%, representing a 12.59% improvement over the original MobileNet-V2. Compared to other lightweight neural networks—such as EfficientNet-B0, MNASNet, MobileNet-V2, ResNet-18, ShuffleNet-V2, and SqueezeNet—CBAM-MobileNet-V2 demonstrates superior recognition accuracy while maintaining a lower parameter count, highlighting its efficiency and suitability for practical applications. The deep learning-based recognition system successfully identifies seven common wheat diseases with high accuracy. Its offline, real-time detection capability eliminates the need for cloud deployment, thereby reducing user costs and supporting intelligent wheat disease diagnosis in resource-limited environments.

Key words: wheat disease, image recognition, deep learning, lightweight models, attention mechanism, mobile deployment

Fig. 1

Modeling process of the mobile wheat disease intelligent recognition system"

Fig. 2

Sample images from the wheat disease dataset"

Table 1

Sample number of the dataset"

类别名称
Category name
增强前
Before being
augmented
增强后After being augmented 总数
Total
训练集
Training set
验证集
Validation set
测试集
Test set
健康叶Healthy leaf 5750 4024 1150 576 5750
条锈病Yellow rust 6781 4746 1356 679 6781
叶锈病Brown rust 6169 4318 1234 617 6169
秆锈病Black rust 1594 4463 1275 638 6376
叶枯病Leaf blight 2556 5367 1534 767 7668
白粉病Powdery mildew 1496 4188 1179 599 5966
赤霉病Fusarium head blight 2341 4916 1404 703 7023
散黑穗病Loose smut 2115 4441 1269 635 6345

Fig. 3

Structure diagram of the CBAM-MobileNet-V2 model Conv: convolution; Sigmoid: activation function; CBAM: convolutional block attention module."

Table 2

Training and testing lossand accuracy undertwo different learning strategies"

模型类型
Model type
学习方式
Learning methods
训练损失
Training loss
训练准确率
Training accuracy (%)
测试损失
Test loss
测试准确率
Test accuracy (%)
参数量(×106)
Number of parameters (×106)
模型大小
Model size (Mb)
EfficientNet-B0 全新学习
Brand new learning
0.0203 78.08 0.0154 83.44 4.02 15.33
迁移学习
Transfer learning
0.0010 98.73 0.0073 94.62 4.02 15.33
MNASNet 全新学习
Brand new learning
0.0182 80.19 0.0140 85.42 3.11 11.87
迁移学习
Transfer learning
0.0006 99.18 0.0111 92.87 3.11 11.87
MobileNet-V2 全新学习
Brand new learning
0.0177 81.03 0.0142 85.36 2.23 8.52
迁移学习
Transfer learning
0.0006 99.19 0.0114 92.43 2.23 8.52
ResNet-18 全新学习
Brand new learning
0.0184 80.04 0.0175 81.14 11.18 42.65
迁移学习
Transfer learning
0.0004 99.30 0.0132 91.98 11.18 42.65
ShuffleNet-V2 全新学习
Brand new learning
0.0325 70.19 0.0309 71.15 1.26 4.81
迁移学习
Transfer learning
0.0006 99.23 0.0138 91.32 1.26 4.81
SqueezeNet 全新学习
Brand new learning
0.0174 81.47 0.0168 82.21 0.74 2.82
迁移学习
Transfer learning
0.0034 96.72 0.0157 88.58 0.74 2.82

Table 3

Performance impact of different downsampling methods"

模型
Model
采样方式
Sampling method
训练损失
Training loss
训练准确率
Training accuracy (%)
测试损失
Test loss
测试准确率
Test accuracy (%)
MobileNet-V2 平均池化层 0.0006 99.19 0.0114 92.43
Average pooling layer
最大池化层 0.0006 99.20 0.0113 92.20
Maximum pooling layer
最小池化层 0.0011 98.70 0.0129 91.77
Minimum pooling layer

Table 4

Performance impactof different attention mechanisms"

模型
Model
注意力机制
Attention mechanism
训练损失
Training loss
训练准确率
Training
accuracy (%)
测试
损失
Test loss
测试准确率
Test accuracy
(%)
参数量(×106)
Number of
parameters (×106)
模型大小
Model size
(Mb)
MobileNet-V2 CBAM 0.0006 99.15 0.0109 93.00 2.54 9.69
CA 0.0007 99.15 0.0119 92.31 2.54 9.69
ECA 0.0006 99.15 0.0107 92.72 7.15 27.27
SE 0.0006 99.17 0.0106 92.73 2.44 9.30
SimAM 0.0007 99.04 0.0114 92.22 5.51 21.03

Fig. 4

Visualization of heatmaps for different attention modules"

Table 5

Effects of various regularization coefficients and learning rates on model performance"

模型
Model
正则项系数
Regularization coefficient
学习率
Learning rate
训练损失
Training loss
训练准确率
Training accuracy (%)
测试损失
Test loss
测试准确率
Test accuracy (%)
MobileNet-V2 0 0.001 0.0007 99.14 0.0109 93.00
0.005 0.0006 99.15 0.0062 95.37
0.010 0.0004 99.28 0.0061 96.04
0.001 0.001 0.0007 99.10 0.0060 93.99
0.005 0.0006 99.20 0.0052 95.56
0.010 0.0005 99.22 0.0050 96.10
0.005 0.001 0.0039 96.76 0.0062 95.35
0.005 0.0014 98.98 0.0051 96.04
0.010 0.0008 99.17 0.0048 96.13

Table 6

Effects of different optimizers on model performance"

模型
Model
优化器
Optimizer
训练损失
Training loss
训练准确率
Training accuracy (%)
测试损失
Test loss
测试准确率
Test accuracy (%)
MobileNet-V2 SGD 0.0008 99.17 0.0048 96.13
Adam 0.0004 99.33 0.0097 95.49
AdamW 0.0004 99.28 0.0084 97.95
CAME 0.0081 91.65 0.0081 91.64
LBFGS 2.1001 10.66 2.0985 9.38

Table 7

Ablation study of the MobileNet-V2 model"

训练策略
Training strategy
训练损失
Training loss
训练准确率
Training
accuracy (%)
测试损失
Test loss
测试准确率
Test accuracy (%)
参数量(×106)
Number of
parameters (×106)
模型大小
Model size (Mb)
MobileNet-V2 0.0177 81.03 0.0142 85.36 2.23 8.52
MobileNet-V2+全局平均池化
MobileNet-V2+global average pooling layer
0.0006 99.19 0.0114 92.43 2.23 8.52
MobileNet-V2+全局平均池化+CBAM
MobileNet-V2+global average pooling layer+CBAM
0.0006 99.15 0.0109 93.00 2.54 9.69
MobileNet-V2+全局平均池化+
CBAM+0.005+0.01
MobileNet-V2+global average pooling layer+CBAM+0.005+0.01
0.0008 99.17 0.0048 96.13 2.54 9.69
MobileNet-V2+全局平均池化+
CBAM+0.005+0.01+AdamW
MobileNet-V2+global average pooling layer+CBAM+0.005+0.01+AdamW
0.0004 99.28 0.0084 97.95 2.54 9.69

Fig. 5

Accuracy and loss curves of different models"

Table 8

Classification results of the proposed model (%)"

病害类别
Disease category
准确率
Accuracy
精确度
Precision
召回率
Recall
F1分数
F1 score
健康叶Healthy leaf 97.1 95.9 96.6 96.2
散黑穗病Loose smut 97.3 97.6 95.8 96.7
秆锈病Black rust 87.9 87.5 89.5 88.5
叶锈病Brown rust 97.6 90.8 92.9 91.8
叶枯病 Leaf blight 93.0 94.2 91.4 92.8
白粉病Powdery mildew 98.3 97.0 96.7 96.9
赤霉病Fusarium head blight 97.0 96.1 95.4 95.7
条锈病Yellow rust 81.9 89.8 91.0 90.4

Fig. 6

Confusion matrix of the MobileNet-V2 model"

Fig. 7

Development process of the wheat disease recognition mobile application"

Fig. 8

ONNX format conversion flowchart"

Fig. 9

Userinterface of the wheat disease recognition app on mobile devices Left: Redmi K70E (MediaTek platform, Dimensity 8300 Ultra); Middle: OnePlus ace2 (Qualcomm platform, Snapdragon 8+Gen 1); Right: Huawei nova5pro (HarmonyOS platform, Kirin 980)."

Fig. 10

Comprehensive chip performance ranking by GeekBay The performance test is scored based on the Snapdragon 865 (100 points). Among them, the CPU performance accounts for 70% and the GPU performance accounts for 30%."

[1] 刘志勇, 王道文, 张爱民, 等. 小麦育种行业创新现状与发展趋势. 植物遗传资源学报, 2018, 19: 430-434.
Liu Z Y, Wang D W, Zhang A M, et al. Current status and perspective of wheat genomics, genetics, and breeding. J Plant Genet Resour, 2018, 19: 430-434 (in Chinese with English abstract).
[2] 孙新华, 尹爱玲. 小麦主要病害及综合防控技术. 种子科技, 2022, 40(6): 99-101.
Sun X H, Yin A L. Major diseases of wheat and comprehensive prevention and control technology. Seed Sci Technol, 2022, 40(6): 99-101 (in Chinese with English abstract).
[3] 彭玉寒, 李书琴. 基于重参数化MobileNetV2的农作物叶片病害识别模型. 农业工程学报, 2023, 39(17): 132-140.
Peng Y H, Li S Q. Recognizing crop leaf diseases using reparameterized MobileNetV2. Trans CSAE, 2023, 39(17): 132-140 (in Chinese with English abstract).
[4] Nigam S, Jain R, Marwaha S, et al. Deep transfer learning model for disease identification in wheat crop. Ecol Inform, 2023, 75: 102068.
doi: 10.1016/j.ecoinf.2023.102068
[5] 何东健, 王鹏, 牛童, 等. 基于改进残差网络的田间葡萄霜霉病病害程度分级模型. 农业机械学报, 2022, 53(1): 235-243.
He D J, Wang P, Niu T, et al. Classification model of grape downy mildew disease degree in field based on improved residual network. Trans CSAM, 2022, 53(1): 235-243 (in Chinese with English abstract).
[6] 王守志, 何东健, 李文, 等. 基于核K-均值聚类算法的植物叶部病害识别. 农业机械学报, 2009, 40(3): 152-155.
Wang S Z, He D J, Li W, et al. Plant leaf disease recognition based on kernel K-means clustering algorithm. Trans CSAM, 2009, 40(3): 152-155 (in Chinese with English abstract).
[7] 秦立峰, 张延苏. 混合颜色特征下番茄叶霉病病斑双层K-means聚类分割方法. 安徽农业科学, 2018, 46(3): 169-170.
Qin L F, Zhang Y S. Double-layer K-means clustering segmentation method of tomato leaf mold disease spots with mixed color features. J Anhui Agric Sci, 2018, 46(3): 169-170 (in Chinese with English abstract).
[8] 叶海建, 郎睿, 刘成启, 等. 基于视觉显著性图的黄瓜霜霉病识别方法. 农业机械学报, 2016, 47(5): 270-274.
Ye H J, Lang R, Liu C Q, et al. Recognition of cucumber downy mildew disease based on visual saliency map. Trans CSAM, 2016, 47(5): 270-274 (in Chinese with English abstract).
[9] 陈荣, 李旺, 周文玉. 基于形状特征和支持向量机(SVM)的茶叶病害识别方法. 贵州农业科学, 2021, 49(4): 53-59.
Chen R, Li W, Zhou W Y. Recognition method of tea disease based on shape features and support vector machine (SVM). Guizhou Agric Sci, 2021, 49(4): 53-59 (in Chinese with English abstract).
[10] 李亚文, 何甜. 基于GA-SVM和特征提取的苹果叶部病害识别检测. 食品与发酵科技, 2024, 60(3): 7-13.
Li Y W, He T. Detection of apple leaf disease identification based on GA-SVM and feature extraction. Food Ferment Sci Technol, 2024, 60(3): 7-13 (in Chinese with English abstract).
[11] Xu P F, Wu G S, Guo Y J, et al. Automatic wheat leaf rust detection and grading diagnosis via embedded image processing system. Procedia Comput Sci, 2017, 107: 836-841.
doi: 10.1016/j.procs.2017.03.177
[12] 李冠林, 马占鸿, 王海光. 基于支持向量机的小麦条锈病和叶锈病图像识别. 中国农业大学学报, 2012, 17(2): 72-79.
Li G L, Ma Z H, Wang H G. Image recognition of wheat stripe rust and wheat leaf rust based on support vector machine. J China Agric Univ, 2012, 17(2): 72-79 (in Chinese with English abstract).
[13] 郭青, 王骊雯, 董方敏, 等. 基于方向一致性特征的小麦条锈病与白粉病识别方法. 农业机械学报, 2015, 46(1): 26-34.
Guo Q, Wang L W, Dong F M, et al. Identification of wheat stripe rust and powdery mildew using orientation coherence feature. Trans CSAM, 2015, 46(1): 26-34 (in Chinese with English abstract).
[14] 张云龙, 袁浩, 张晴晴, 等. 基于颜色特征和差直方图的苹果叶部病害识别方法. 江苏农业科学, 2017, 45(14): 171-174.
Zhang Y L, Yuan H, Zhang Q Q, et al. Identification method of apple leaf disease based on color characteristics and differential histograms. Jiangsu Agric Sci, 2017, 45(14): 171-174 (in Chinese with English abstract).
[15] 刘梦姝, 张春琪, 晁金阳, 等. 基于YOLO v8n改进的小麦病害检测系统. 农业机械学报, 2024, 55(增刊1): 280-287.
Liu M S, Zhang C Q, Chao J Y, et al. Improved wheat disease detection system based on YOLO v8n. Trans CSAM, 2024, 55(S1): 280-287 (in Chinese with English abstract).
[16] 杜海顺, 张春海, 安文昊, 等. 基于多层信息融合和显著性特征增强的农作物病害识别. 农业机械学报, 2023, 54(7): 214-222.
Du H S, Zhang C H, An W H, et al. Crop disease recognition based on multi-layer information fusion and saliency feature enhancement. Trans CSAM, 2023, 54(7): 214-222 (in Chinese with English abstract).
[17] Praveen S P, Nakka R, Chokka A, et al. A novel classification approach for grape leaf disease detection based on different attention deep learning techniques. Int J Adv Comput Sci Appl, 2023, 14: 1199-1209.
[18] 龙满生, 欧阳春娟, 刘欢, 等. 基于卷积神经网络与迁移学习的油茶病害图像识别. 农业工程学报, 2018, 34(18): 194-201.
Long M S, Ou-Yang C J, Liu H, et al. Image recognition of Camellia oleifera diseases based on convolutional neural network & transfer learning. Trans CSAE, 2018, 34(18): 194-201 (in Chinese with English abstract).
[19] Zhong Y, Zhao M. Research on deep learning in apple leaf disease recognition. Comput Electron Agric, 2020, 168: 105146.
doi: 10.1016/j.compag.2019.105146
[20] 郭旭超, 唐詹, 刁磊, 等. 基于部首嵌入和注意力机制的病虫害命名实体识别. 农业机械学报, 2020, 51(增刊2): 335-343.
Guo X C, Tang Z, Diao L, et al. Recognition of Chinese agricultural diseases and pests named entity with joint radical-embedding and self-attention mechanism. Trans CSAM, 2020, 51(S2): 335-343 (in Chinese with English abstract).
[21] 许景辉, 邵明烨, 王一琛, 等. 基于迁移学习的卷积神经网络玉米病害图像识别. 农业机械学报, 2020, 51(2): 230-236.
Xu J H, Shao M Y, Wang Y C, et al. Image recognition of corn diseases based on convolutional neural network and transfer learning. Trans CSAM, 2020, 51(2): 230-236 (in Chinese with English abstract).
[22] Mohanty S P, Hughes D P, Salathé M. Using deep learning for image-based plant disease detection. Front Plant Sci, 2016, 7: 1419.
pmid: 27713752
[23] Sladojevic S, Arsenovic M, Anderla A, et al. Deep neural networks based recognition of plant diseases by leaf image classification. Comput Intell Neurosci, 2016, 2016: 3289801.
[24] Brahimi M, Boukhalfa K, Moussaoui A. Deep learning for tomato diseases: classification and symptoms visualization. Appl Artif Intell, 2017, 31: 299-315.
doi: 10.1080/08839514.2017.1315516
[25] 张建华, 孔繁涛, 吴建寨, 等. 基于改进VGG卷积神经网络的棉花病害识别模型. 中国农业大学学报, 2018, 23(11): 161-171.
Zhang J H, Kong F T, Wu J Z, et al. Cotton disease identification model based on improved VGG convolution neural network. J China Agric Univ, 2018, 23(11): 161-171 (in Chinese with English abstract).
[26] Chen J D, Zhang D F, Suzauddola M, et al. Identifying crop diseases using attention embedded MobileNet-V2 model. Appl Soft Comput, 2021, 113: 107901.
doi: 10.1016/j.asoc.2021.107901
[27] Lu L, Liu W, Yang W B, et al. Lightweight corn seed disease identification method based on improved ShuffleNetV2. Agriculture, 2022, 12: 1929.
doi: 10.3390/agriculture12111929
[28] 刘洋, 冯全, 王书志. 基于轻量级CNN的植物病害识别方法及移动端应用. 农业工程学报, 2019, 35(17): 194-204.
Liu Y, Feng Q, Wang S Z. Plant disease identification method based on lightweight CNN and mobile application. Trans CSAE, 2019, 35(17): 194-204 (in Chinese with English abstract).
[29] 俞佩仕, 郭龙军, 姚青, 等. 基于移动终端的稻田飞虱调查方法. 昆虫学报, 2019, 62: 615-623.
Yu P S, Guo L J, Yao Q, et al. Investigation method of rice planthopper based on mobile terminal. Acta Entomol Sin, 2019, 62: 615-623 (in Chinese with English abstract).
[30] 胡灿, 王兴旺, 王旭峰, 等. 基于Android手机的田间棉花产量预测系统设计. 农业机械学报, 2023, 54(增刊2): 252-259.
Hu C, Wang X W, Wang X F, et al. Design of field cotton yield forecasting system based on Android mobile phone. Trans CSAM, 2023, 54(S2): 252-259 (in Chinese with English abstract).
[31] 郑鹏辉.基于Android平台的温室蔬菜病害程度检测系统研究. 沈阳农业大学硕士学位论文, 辽宁沈阳, 2016.
Zheng P H. Research on the Detection System of the Greenhouse Vegetable Diseases Degree Based on Android Platform. MS Thesis of Shenyang Agricultural University, Shenyang, Liaoning, China, 2016 (in Chinese with English abstract).
[32] 乐羿成.基于Android的作物冠层抓拍传输及其图像自动管理系统开发. 浙江理工大学硕士学位论文, 浙江杭州, 2017.
Le Y C. Development of Android Based Capture & Transmission of Crop Canopy and Automatic System on Its Image Managment. MS Thesis of Zhejiang Sci-Tech University, Hangzhou, Zhejiang, China, 2017 (in Chinese with English abstract).
[33] 庄福振, 罗平, 何清, 等. 迁移学习研究进展. 软件学报, 2015, 26(1): 26-39.
Zhuang F Z, Luo P, He Q, et al. Survey on transfer learning research. J Softw, 2015, 26(1): 26-39 (in Chinese with English abstract).
[34] Luo Y, Ren X Z, Zheng Z W, et al. CAME: confidence-guided adaptive memory efficient optimization. 2023: arXiv: 2307.02047. https://arxiv.org/abs/2307.02047.
[35] 胡施威, 邱林, 邓建新. 基于改进MobileViT的葡萄叶部病害识别模型. 山东农业科学, 2024, 56(10): 159-166.
Hu S W, Qiu L, Deng J X. Grape leaf disease recognition model based on improved MobileViT. Shandong Agric Sci, 2024, 56(10): 159-166 (in Chinese with English abstract).
[36] 彭红星, 徐慧明, 刘华鼐. 融合双分支特征和注意力机制的葡萄病虫害识别模型. 农业工程学报, 2022, 38(10): 156-165.
Peng H X, Xu H M, Liu H N. Model for identifying grape pests and diseases based on two-branch feature fusion and attention mechanism. Trans CSAE, 2022, 38(10): 156-165 (in Chinese with English abstract).
[37] 王欣, 王美丽, 边党伟. 融合MobileNetv2和注意力机制的轻量级人像分割算法. 计算机工程与应用, 2022, 58(7): 220-228.
doi: 10.3778/j.issn.1002-8331.2106-0334
Wang X, Wang M L, Bian D W. Algorithm for portrait segmentation combined with MobileNetv2 and attention mechanism. Comput Eng Appl, 2022, 58(7): 220-228 (in Chinese with English abstract).
doi: 10.3778/j.issn.1002-8331.2106-0334
[38] 牛思琪, 马睿, 许晓琳, 等. 基于改进CBAM注意力机制的MobileNetV2玉米种子品种识别研究. 中国粮油学报, 2024, 39(3): 159-165.
Niu S Q, Ma R, Xu X L, et al. Mobile NetV2 maize seed variety recognition based on improved attention mechanism CBAM. J Chin Cereals Oils Assoc, 2024, 39(3): 159-165 (in Chinese with English abstract).
[39] Loshchilov I, Hutter F. Decoupled weight decay regularization. International Conference on Learning Representations, 2017.
[40] 何汝艳, 吕子俊, 贾森. 基于LBFGS加速多层感知网络的高光谱遥感非光学水质参数反演. 遥感学报, 2024, 28: 2970-2983.
He R Y, Lyu Z J, Jia S. Inversion of non-optical water quality parameters of hyperspectral remote sensing based on LBFGS-accelerated multi-layer perceptron network. Natl Remote Sens Bull, 2024, 28: 2970-2983 (in Chinese with English abstract).
[41] 周罕觅, 陈佳庚, 代智光, 等. 基于知识蒸馏和轻量级卷积神经网络的植物病害识别方法. 南京农业大学学报, 2024, 47: 1189-1201.
Zhou H M, Chen J G, Dai Z G, et al. Plant leaf disease recognition based on knowledge distillation and lightweight convolutional neural network. J Nanjing Agric Univ, 2024, 47: 1189-1201 (in Chinese with English abstract).
[42] Hou Q B, Zhou D Q, Feng J S. Coordinate attention for efficient mobile network design. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). June 20-25, 2021. Nashville, TN, USA. IEEE, 2021: 13708-13717.
[43] 袁培森, 欧阳柳江, 翟肇裕, 等. 基于MobileNetV3Small-ECA的水稻病害轻量级识别研究. 农业机械学报, 2024, 55(1): 253-262.
Yuan P S, Ou-Yang L J, Zhai Z Y, et al. Lightweight identification of rice diseases based on improved ECA and MobileNetV3Small. Trans CSAM, 2024, 55(1): 253-262 (in Chinese with English abstract).
[44] 李子茂, 徐杰, 郑禄, 等. 基于改进DenseNet的茶叶病害小样本识别方法. 农业工程学报, 2022, 38(10): 182-190.
Li Z M, Xu J, Zheng L, et al. Small sample recognition method of tea disease based on improved DenseNet. Trans CSAE, 2022, 38(10): 182-190 (in Chinese with English abstract).
[45] 田甜, 程志友, 鞠薇, 等. 基于SimAM-ConvNeXt-FL的茶叶病害小样本分类方法研究. 农业机械学报, 2024, 55(3): 275-281.
Tian T, Cheng Z Y, Ju W, et al. Small sample classification of tea diseases based on SimAM-ConvNeXt-FL. Trans CSAE, 2024, 55(3): 275-281 (in Chinese with English abstract).
[46] Xu F. Faster scalable ML model deployment using ONNX and open source tools. Proceedings of the 2020 IEEE Infrastructure Conference, 2020: 9377608.
[47] 陈博远.基于移动端的羊脸身份识别模型设计和软件开发. 内蒙古农业大学硕士学位论文, 内蒙古呼和浩特, 2024.
Chen B Y. Design and Software Development of Sheep Face Identification Model Based on Mobile Terminal. MS Thesis of Inner Mongolia Agricultural University, Hohhot, Inner Mongolia, China, 2024 (in Chinese with English abstract).
[48] 邹小波, 高文健, 石吉勇, 等. 基于改进DenseNet与迁移学习的食品图像分类技术. 中国农机化学报, 2025, 46(6): 77-84.
Zou X B, Gao W J, Shi J Y, et al. Food image classification technology based on improved DenseNet and transfer learning. J Chin Agric Mech, 2025, 46(6): 77-84 (in Chinese with English abstract).
[49] 王中璞, 吴正香, 尤美路, 等. 基于改进MobileNetV2的棉花颜色分级检测. 棉纺织技术, 2024, 52(6): 15-21.
Wang Z P, Wu Z X, You M L, et al. Cotton color grade detection based on improved MobileNetV2. Cotton Text Technol, 2024, 52(6): 15-21 (in Chinese with English abstract).
[50] 任治洲, 梁琨, 王泽宇, 等. 结合改进CBAM和MobileNetV2算法对小麦病斑粒分类. 南京农业大学学报, 2024, 47: 583-591.
Ren Z Z, Liang K, Wang Z Y, et al. Combining improved CBAM and MobileNetV2 algorithms for classifying diseased wheat kernels. J Nanjing Agric Univ, 2024, 47: 583-591 (in Chinese with English abstract).
[51] 王发明, 倪昕东, 张旗, 等. 基于MobileNetV2-CBAM的机收场景下冬小麦成熟期在线分类识别方法. 农业机械学报, 2024, 55(增刊1): 71-80.
Wang F M, Ni X D, Zhang Q, et al. Online classification and identification method for winter wheat at maturity stage in combine harvesting scenarios based on MobileNetV2-CBAM. Trans CSAM, 2024, 55(S1): 71-80 (in Chinese with English abstract).
[52] 邵泽中, 姚青, 唐健, 等. 面向移动终端的农业害虫图像智能识别系统的研究与开发. 中国农业科学, 2020, 53: 3257-3268.
doi: 10.3864/j.issn.0578-1752.2020.16.005
Shao Z Z, Yao Q, Tang J, et al. Research and development of the intelligent identification system of agricultural pests for mobile terminals. Sci Agric Sin, 2020, 53: 3257-3268 (in Chinese with English abstract).
doi: 10.3864/j.issn.0578-1752.2020.16.005
[1] LI Jin-Min, CHEN Xiu-Qing, YANG Qi, SHI Liang-Sheng. Deep learning models for estimation of paddy rice leaf nitrogen concentration based on canopy hyperspectral data [J]. Acta Agronomica Sinica, 2021, 47(7): 1342-1350.
[2] YAN Zhuang-Zhuang, YAN Xue-Hui, SHI Jia, SUN Kai, YU Jiang-Lin, ZHANG Zhan-Guo, HU Zhen-Bang, JIANG Hong-Wei, XIN Da-Wei, LI Yang, QI Zhao-Ming, LIU Chun-Yan, WU Xiao-Xia, CHEN Qing-Shan, ZHU Rong-Sheng. Classification of soybean pods using deep learning [J]. Acta Agronomica Sinica, 2020, 46(11): 1771-1779.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!