簡易檢索 / 詳目顯示
| 研究生： |
李湘渝 Li, Xiang-Yu |
|---|---|
| 論文名稱： |
基於卷積神經網路之深度學習方法之蘋果葉面病害辨識與分類 Identification and Classification of Apple Leaf Diseases by Deep Learning Method based on Convolutional Neural Network |
| 指導教授： |
劉書助
Liu, Shu-Chu |
| 學位類別： |
碩士 Master |
| 系所名稱： |
管理學院 - 資訊管理系所 Department of Management Information Systems |
| 論文出版年： | 2022 |
| 畢業學年度： | 111 |
| 語文別： | 中文 |
| 論文頁數： | 41 |
| 中文關鍵詞： | 深度學習 、ResNet-50 、卷積神經網路 、DenseNet-201 、蘋果病害辨識與分類 、MobileNet 、智慧農業 、EfficientNet-B0 |
| 外文關鍵詞： | Deep Learning, ResNet-50, Convolutional Neural Network, DenseNet-201, Apple Disease Identification and Classification, MobileNet, Smart Agriculture, EfficientNet-B0 |
| DOI URL： | http://doi.org/10.6346/NPUST202200458 |
| 相關次數： | 點閱：143 下載：15 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 |
農業的生產力與競爭力對於國家經濟發展的穩定性與持久性有著直接的關係，智慧農業指的是傳統農業結合影像辨識、物聯網、大數據分析等前瞻性技術，達到增產增量、精準用藥、災變預警等智能管理的一種方法。在蘋果的種植過程中，作物病蟲害是決定蘋果產量和質量的重要因素，錯誤和延遲的診斷可能導致過度或不充分地使用化學品，從而增加生產成本、環境負擔及身體健康的憂慮，因此，作物疾病管理的適當和及時部署是至關重要的。在本文中，我們利用 ResNet-50、DenseNet-201、MobileNet、EfficientNet-B0 四種卷積神經網路 (Convolution Neural Network, CNN) 方法，用以辨識和分類蘋果葉片病害，並將其病害分為健康、蘋果黑星病、銹病和多種疾病四種不同類型，最後再比較這四種深度學習方法，使用的數據集是國外公開競賽平台 Kaggle 用於區分蘋果葉面疾病的圖像，本研究使用經過篩選過後的 2,688 張實時圖像，對蘋果葉面疾病進行辨識與分類，根據實驗結果顯示 EfficientNet-B0 相較於其他三種模型，它對於辨別與分類蘋果葉片病害的表現最為優秀，能夠準確的識別受感染的作物葉面，並對其疾病類型進行分類，此方法在診斷健康、蘋果黑星病、銹病和多種疾病這四種不同類型疾病所獲得的 F1-Score 分別為 99.66%、99.20%、99.53%、90.82%；在準確率方面，ResNet-50、DenseNet-201、MobileNet、EfficientNet-B0 四種模型所獲得的分數為 98.83%、98.61%、94.49%、99.15%。本研究期望透過提供有關受感染的作物信息來提醒及幫助農民提前預防及阻止病害之蔓延，並有效提升蘋果的生產力，與提高蘋果的質量，和降低勞動力成本。
The productivity and competitiveness of agriculture are directly related to the stability and durability of a country's economic development. Smart agriculture refers to the combination of traditional agriculture with forward-looking technologies such as image recognition, Internet of Things, and big data analysis to achieve intelligent management such as increased production, precision medicine, and disaster warning. In the process of apple planting, crop diseases and insect pests are important factors that determine the yield and quality of apples. if incorrect and delayed diagnosis happens, it may lead to excessive or insufficient use of chemicals. Further, it might increase production costs, environmental burdens as well as health concerns. Thus, accurate crop disease management and preventive measures are critical. Four Convolution Neural Network (CNN) methods (ResNet-50, DenseNet-201, MobileNet, EfficientNet-B0) were designed to identify and classify apple leaf diseases. The diseases can be divided into four different types: health, apple scab, rust and multiple diseases. The aims are to evaluate the four deep learning methods. The images source of the data set is provided by Kaggle, a foreign open competition platform. This study uses 2,688 screened real-time images to identify and classify apple leaf diseases. The results show that EfficientNet-B0 has the best performance in identifying and classifying apple leaf diseases compared with the other three models. The F1-Scores for EfficientNet-B0 are 99.66%, 99.20%, 99.53% and 90.82% in diagnosing four different types of diseases: health, apple scab, rust, and multiple diseases. In terms of accuracy, the scores obtained by the four models of ResNet-50, DenseNet-201, MobileNet, and EfficientNet-B0 are 98.83%, 98.61%, 94.49%, and 99.15%. This study provides the information about infected crops to prevent the spread of the disease, improve the productivity of apples, and reduce labor costs.
第1章 緒論 1
1.1 研究背景 1
1.2 問題與動機 2
1.3 研究目的 3
1.4 論文架構 4
第2章 文獻探討 5
2.1 卷積神經網路在影像辨識與分類上的應用 5
2.2 RESNET 模型 8
2.3 DENSENET 模型 9
2.4 MOBILENET 模型 10
2.5 EFFICIENTNET 模型 11
第3章 研究方法 12
3.1 資料集 12
3.2 模型架構 14
3.3 RESNET-50 模型 17
3.4 DENSENET-201 模型 18
3.5 MOBILENET 模型 19
3.6 EFFICIENTNET-B0 模型 21
3.7 模型評估 22
第4章 實驗結果 24
4.1 實驗環境 24
4.2 RESNET-50實驗結果 25
4.3 DENSENET-201實驗結果 27
4.4 MOBILENET實驗結果 29
4.5 EFFICIENTNET-B0實驗結果 31
4.6 四種模型的實驗結果比較 33
4.7 四種模型的準確率 35
第5章 結論與未來展望 36
5.1 結論 36
5.2 研究限制與未來展望 37
參考文獻 38
[1] Barman, U., Choudhury, R. D., Sahuc, D., & Barman, G. G. (2020). Comparison of Convolution Neural Networks for Smartphone Image Based Real Time Classification of Citrus Leaf Disease. Computers and Electronics in Agriculture, 177, 105661. https://doi.org/10.1016/j.compag.2020.105661.
[2] Bhupendra, Moses, K., Miglani, A., & Kankar, P. K. (2022). Deep CNN-Based Damage Classification of Milled Rice Grains Using a High-Magnification Image Dataset. Computers and Electronics in Agriculture, 195, 106811. https://doi.org/10.1016/j.compag.2022.106811.
[3] Bi, C., Wang, J., Duan, Y., Fu, B., Kang, R. J., & Shi, Y. (2022). MobileNet Based Apple Leaf Diseases Identification. Mobile Networks and Applications, 27, 172–180 https://doi.org/10.1007/s11036-020-01640-1.
[4] Bowen, J. K., Mesarich, C. H., Bus, V. G. M., Beresford, R. M., Plummer, K. M., & Templeton, M. D. (2010). Venturia Inaequalis : The Causal Agent of Apple Scab. Molecular Plant Pathology, 12, 2, 105-122 https://doi.org/10.1111/j.1364-3703.2010.00656.x.
[5] Cetinoglu, Y. K., Koska, I. O., Uluc, M. E., & Gelal, M. F. (2021). Detection and Vascular Territorial Classification of Stroke on Diffusion-Weighted MRI by Deep Learning. European Journal of Radiology, 145, 110050. https://doi.org/10.1016/j.ejrad.2021.110050.
[6] Deshpande, A., Estrela, V. V., & Patavardhan, P. (2021). The DCT-CNN-ResNet-50 Architecture to Classify Brain Tumors with Super-Resolution, Convolutional Neural Network, and the ResNet-50. Neuroscience Informatics, 1, 4, 100013. https://doi.org/10.1016/j.neuri.2021.100013.
[7] Emanuel, I. B., Ralston, T. I., Chatfield, J., Draper, E., Veil, J., & Hand, F. P. (2021). First Report of Gymnosporangium Yamadae Causing Japanese Apple Rust on Crabapple (Malus Spp.) in Ohio. The American Phytopathological Society, 105, 7. https://doi.org/10.1094/PDIS-12-20-2612-PDN.
[8] Fenu, G., & Malloci, F. M. (2021). DiaMOS Plant: A Dataset for Diagnosis and Monitoring Plant Disease. Agronomy, 11, 11. https://doi.org/10.3390/agronomy11112107.
[9] He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770-778. https://doi.org/10.1109/CVPR.2016.90.
[10] Hossain, M. B., Iqbal, S. M. H. S., Islam, M. M., Akhtar, M. N., & Sarker, I. H. (2022). Transfer Learning with Fine-Tuned Deep CNN ResNet-50 Model for Classifying COVID-19 from Chest X-Ray Images. Informatics in Medicine Unlocked, 30, 100916. https://doi.org/10.1016/j.imu.2022.100916.
[11] Howard, A. G., Zhu, M., Chen, B., Kalenichenko, D., Wang, W., Weyand, T., Andreetto, M., & Adam, H. (2017). MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications. arXiv.1704.04861. https://doi.org/10.48550/arXiv.1704.04861.
[12] Huang, G., Liu, Z., Maaten, L. V., & Weinberger, K. Q. (2017). Densely Connected Convolutional Networks. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2261-2269. https://doi.org/10.1109/CVPR.2017.243.
[13] Liew, X. Y., Hameed, N., & Clos, J. (2021). An Investigation of XGBoost-Based Algorithm for Breast Cancer Classification. Machine Learning with Applications, 6, 100154. https://doi.org/10.1016/j.mlwa.2021.100154.
[14] Liu, B., Zhang, Y., He, D., & Li, Y. (2017). Identification of Apple Leaf Diseases Based on Deep Convolutional Neural Networks. Symmetry, 10, 1. https://doi.org/10.3390/sym10010011.
[15] Liu, P., Wang, Y., Meng, J., Zhang, X., Zhou, J., Han, M., Yang, C., Gan, L., & Li , H. (2019). Transcriptome Sequencing and Expression Analysis of Genes Related to Anthocyanin Biosynthesis in Leaves of Malus ‘Profusion’ Infected by Japanese Apple Rust. Forests, 10, 8. https://doi.org/10.3390/f10080665.
[16] Lu, S. Y., Wang, S. H., & Zhang, Y. D. (2020). A Classification Method for Brain MRI via MobileNet and Feedforward Network with Random Weights. Pattern Recognition Letters, 140, 252–260. https://doi.org/10.1016/j.patrec.2020.10.017.
[17] Lu, Y., Chen, Q., Bu, Y., Luo, R., Hao, S., Zhang, J., Tian, J., & Yao, Y. (2017). Flavonoid Accumulation Plays an Important Role in the Rust Resistance of Malus Plant Leaves. Front. Plant Science, 8, 1286. https://doi.org/10.3389/fpls.2017.01286.
[18] Park, K., Hong, Y. K., Kim, G. H., & Lee, J. (2018). Classification of Apple Leaf Conditions in Hyper-Spectral Images for Diagnosis of Marssonina Blotch Using MRMR and Deep Neural Network. Computers and Electronics in Agriculture, 148, 179-187. https://doi.org/10.1016/j.compag.2018.02.025.
[19] Patocchi, A., Bigler, B., Koller, B. Kellerhals, M., & Gessler, C. (2004). Vr2 A New Apple Scab Resistance Gene. Theoretical and Applied Genetics, 109, 1087–1092 https://doi.org/10.1007/s00122-004-1723-8.
[20] Rahman, M., Cao, Y., Sun, X., Li, B., & Hao, Y. (2021). Deep Pre-Trained Networks as a Feature Extractor with XGBoost to Detect Tuberculosis from Chest X-Ray. Computers & Electrical Engineering, 93, 107252. https://doi.org/10.1016/j.compeleceng.2021.107252.
[21] Sibiya, M., & Sumbwanyambe, M. (2021). Automatic Fuzzy Logic-Based Maize Common Rust Disease Severity Predictions with Thresholding and Deep Learning. Pathogens, 10, 2. https://doi.org/10.3390/pathogens10020131.
[22] Singh, K.P. (2019). Aerobiology, Epidemiology and Management Strategies in Apple Scab: Science and Its Applications. Indian Phytopathology, 72, 381–408 https://doi.org/10.1007/s42360-019-00162-5.
[23] Subetha, T., Khilar, R., & Christo, M. S. (2021). A Comparative Analysis on Plant Pathology Classification Using Deep Learning Architecture – Resnet and VGG19. Materialstoday: Proceedings. https://doi.org/10.1016/j.matpr.2020.11.993.
[24] Tan, M., & Le, Q. V. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks. arXiv.1905.11946. https://doi.org/10.48550/arXiv.1905.11946.
[25] Thapa, R., Zhang, K., Snavely, N., Belongie, S., & Khan, A. (2020). The Plant Pathology Challenge 2020 Dataset to Classify Foliar Disease of Apples. Applications in Plant Sciences, 8, 9. https://doi.org/10.1002/aps3.11390.
[26] Tiwari, V., Joshi, R. C., & Dutta, M. K. (2021). Dense Convolutional Neural Networks Based Multiclass Plant Disease Detection and Classification Using Leaf Images. Ecological Informatics, 63, 101289. https://doi.org/10.1016/j.ecoinf.2021.101289.
[27] Xiao, K., Zhou, L., Yang, H., & Yang, L. (2022). Phalaenopsis Growth Phase Classification Using Convolutional Neural Network. Smart Agricultural Technology, 2, 100060. https://doi.org/10.1016/j.atech.2022.100060.
[28] Yadav, A., Thakur, U., Saxena, R., Pal, Vipin, Bhateja, V., & Lin, Jerry Chun Wei. (2022). AFD-Net: Apple Foliar Disease Multi Classification Using Deep Learning on Plant Pathology Dataset. Plant and Soil. https://doi.org/10.1007/s11104-022-05407-3.
[29] Yan, Q., Yang, B., Wang, W., Wang, B., Chen, P., & Zhang, J. (2020). Apple Leaf Diseases Recognition Based on An Improved Convolutional Neural Network. Sensors (Basel), 20, 12, 3535. https://doi.org/10.3390/s20123535.
[30] Zhong, Y., & Zhao, M. (2020). Research on Deep Learning in Apple Leaf Disease Recognition. Computers and Electronics in Agriculture, 168, 105146. https://doi.org/10.1016/j.compag.2019.105146.
