Evaluating Critical Disease Occurrence in Grapevine Leaves using CNN: Use-Case in Eastern Europe. Oprea, C., C., Dragulinescu, A., M., C., Marcu, I., M., & Pirnog, I. 2023 17th International Conference on Engineering of Modern Electric Systems, EMES 2023, Institute of Electrical and Electronics Engineers Inc., 2023. ![Evaluating Critical Disease Occurrence in Grapevine Leaves using CNN: Use-Case in Eastern Europe [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper doi abstract bibtex Convolutional Neural Networks are Deep Learning algorithms for image classification tasks in the Computer Vision area. Their efficiency was previously evaluated in medical areas, engineering fields and construction applications. Under this category, VGG16 and Avert-CNN (a modified VGG16 version) algorithms can perform real-time identification for diseases occurrence in agricultural plants with high accuracy and fast estimation time. Thus, this research addresses the identification of health status of grapevine leaves using specialized classification algorithms on images taken from PlantVillage dataset and images acquired in a vineyard in the South-East of Romania. The outcome of these classifications consists in predictions based on one of the 5 classes for which the convolutional network was trained, along with a prediction accuracy metric. The classes considered in this process correspond to the state of a healthy plant and a set of 4 distinct diseases that can affect the vine: Black rot, Esca, Leaf blight and Powdery mildew. Based on the achieved results, a novel convolutional neural network architecture is proposed to ensure reliable estimates on the disease's probability of occurrence. Its efficiency in reaching over 94 % prediction accuracy is demonstrated compared to the classic VGG16 which leads to 90.21 % data accuracy and surpasses Random Forest and Support Vector Machine algorithms that achieve 72.87% and 85.82% accuracy, respectively.
@article{
title = {Evaluating Critical Disease Occurrence in Grapevine Leaves using CNN: Use-Case in Eastern Europe},
type = {article},
year = {2023},
keywords = {oprea2023evaluatingcriticaldisease},
publisher = {Institute of Electrical and Electronics Engineers Inc.},
id = {382c45fa-abf6-3100-9f8d-38a85ca33118},
created = {2023-10-27T07:47:55.667Z},
accessed = {2023-10-27},
file_attached = {true},
profile_id = {f1f70cad-e32d-3de2-a3c0-be1736cb88be},
group_id = {5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1},
last_modified = {2023-11-23T18:12:06.161Z},
read = {true},
starred = {false},
authored = {false},
confirmed = {false},
hidden = {false},
folder_uuids = {6b565182-74c4-44fd-98cc-10618152e2ae,4dd63b9a-2e78-4e16-b45d-b692d8e0c5c3,bd3c6f2e-3514-47cf-bc42-12db8b9abe45},
private_publication = {false},
abstract = {Convolutional Neural Networks are Deep Learning algorithms for image classification tasks in the Computer Vision area. Their efficiency was previously evaluated in medical areas, engineering fields and construction applications. Under this category, VGG16 and Avert-CNN (a modified VGG16 version) algorithms can perform real-time identification for diseases occurrence in agricultural plants with high accuracy and fast estimation time. Thus, this research addresses the identification of health status of grapevine leaves using specialized classification algorithms on images taken from PlantVillage dataset and images acquired in a vineyard in the South-East of Romania. The outcome of these classifications consists in predictions based on one of the 5 classes for which the convolutional network was trained, along with a prediction accuracy metric. The classes considered in this process correspond to the state of a healthy plant and a set of 4 distinct diseases that can affect the vine: Black rot, Esca, Leaf blight and Powdery mildew. Based on the achieved results, a novel convolutional neural network architecture is proposed to ensure reliable estimates on the disease's probability of occurrence. Its efficiency in reaching over 94 % prediction accuracy is demonstrated compared to the classic VGG16 which leads to 90.21 % data accuracy and surpasses Random Forest and Support Vector Machine algorithms that achieve 72.87% and 85.82% accuracy, respectively.},
bibtype = {article},
author = {Oprea, Cristina Claudia and Dragulinescu, Ana Maria Claudia and Marcu, Ioana Manuela and Pirnog, Ionut},
doi = {10.1109/EMES58375.2023.10171678},
journal = {2023 17th International Conference on Engineering of Modern Electric Systems, EMES 2023}
}
Downloads: 0
{"_id":"gwyryeHKJ8TtcSoWv","bibbaseid":"oprea-dragulinescu-marcu-pirnog-evaluatingcriticaldiseaseoccurrenceingrapevineleavesusingcnnusecaseineasterneurope-2023","author_short":["Oprea, C., C.","Dragulinescu, A., M., C.","Marcu, I., M.","Pirnog, I."],"bibdata":{"title":"Evaluating Critical Disease Occurrence in Grapevine Leaves using CNN: Use-Case in Eastern Europe","type":"article","year":"2023","keywords":"oprea2023evaluatingcriticaldisease","publisher":"Institute of Electrical and Electronics Engineers Inc.","id":"382c45fa-abf6-3100-9f8d-38a85ca33118","created":"2023-10-27T07:47:55.667Z","accessed":"2023-10-27","file_attached":"true","profile_id":"f1f70cad-e32d-3de2-a3c0-be1736cb88be","group_id":"5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1","last_modified":"2023-11-23T18:12:06.161Z","read":"true","starred":false,"authored":false,"confirmed":false,"hidden":false,"folder_uuids":"6b565182-74c4-44fd-98cc-10618152e2ae,4dd63b9a-2e78-4e16-b45d-b692d8e0c5c3,bd3c6f2e-3514-47cf-bc42-12db8b9abe45","private_publication":false,"abstract":"Convolutional Neural Networks are Deep Learning algorithms for image classification tasks in the Computer Vision area. Their efficiency was previously evaluated in medical areas, engineering fields and construction applications. Under this category, VGG16 and Avert-CNN (a modified VGG16 version) algorithms can perform real-time identification for diseases occurrence in agricultural plants with high accuracy and fast estimation time. Thus, this research addresses the identification of health status of grapevine leaves using specialized classification algorithms on images taken from PlantVillage dataset and images acquired in a vineyard in the South-East of Romania. The outcome of these classifications consists in predictions based on one of the 5 classes for which the convolutional network was trained, along with a prediction accuracy metric. The classes considered in this process correspond to the state of a healthy plant and a set of 4 distinct diseases that can affect the vine: Black rot, Esca, Leaf blight and Powdery mildew. Based on the achieved results, a novel convolutional neural network architecture is proposed to ensure reliable estimates on the disease's probability of occurrence. Its efficiency in reaching over 94 % prediction accuracy is demonstrated compared to the classic VGG16 which leads to 90.21 % data accuracy and surpasses Random Forest and Support Vector Machine algorithms that achieve 72.87% and 85.82% accuracy, respectively.","bibtype":"article","author":"Oprea, Cristina Claudia and Dragulinescu, Ana Maria Claudia and Marcu, Ioana Manuela and Pirnog, Ionut","doi":"10.1109/EMES58375.2023.10171678","journal":"2023 17th International Conference on Engineering of Modern Electric Systems, EMES 2023","bibtex":"@article{\n title = {Evaluating Critical Disease Occurrence in Grapevine Leaves using CNN: Use-Case in Eastern Europe},\n type = {article},\n year = {2023},\n keywords = {oprea2023evaluatingcriticaldisease},\n publisher = {Institute of Electrical and Electronics Engineers Inc.},\n id = {382c45fa-abf6-3100-9f8d-38a85ca33118},\n created = {2023-10-27T07:47:55.667Z},\n accessed = {2023-10-27},\n file_attached = {true},\n profile_id = {f1f70cad-e32d-3de2-a3c0-be1736cb88be},\n group_id = {5ec9cc91-a5d6-3de5-82f3-3ef3d98a89c1},\n last_modified = {2023-11-23T18:12:06.161Z},\n read = {true},\n starred = {false},\n authored = {false},\n confirmed = {false},\n hidden = {false},\n folder_uuids = {6b565182-74c4-44fd-98cc-10618152e2ae,4dd63b9a-2e78-4e16-b45d-b692d8e0c5c3,bd3c6f2e-3514-47cf-bc42-12db8b9abe45},\n private_publication = {false},\n abstract = {Convolutional Neural Networks are Deep Learning algorithms for image classification tasks in the Computer Vision area. Their efficiency was previously evaluated in medical areas, engineering fields and construction applications. Under this category, VGG16 and Avert-CNN (a modified VGG16 version) algorithms can perform real-time identification for diseases occurrence in agricultural plants with high accuracy and fast estimation time. Thus, this research addresses the identification of health status of grapevine leaves using specialized classification algorithms on images taken from PlantVillage dataset and images acquired in a vineyard in the South-East of Romania. The outcome of these classifications consists in predictions based on one of the 5 classes for which the convolutional network was trained, along with a prediction accuracy metric. The classes considered in this process correspond to the state of a healthy plant and a set of 4 distinct diseases that can affect the vine: Black rot, Esca, Leaf blight and Powdery mildew. Based on the achieved results, a novel convolutional neural network architecture is proposed to ensure reliable estimates on the disease's probability of occurrence. Its efficiency in reaching over 94 % prediction accuracy is demonstrated compared to the classic VGG16 which leads to 90.21 % data accuracy and surpasses Random Forest and Support Vector Machine algorithms that achieve 72.87% and 85.82% accuracy, respectively.},\n bibtype = {article},\n author = {Oprea, Cristina Claudia and Dragulinescu, Ana Maria Claudia and Marcu, Ioana Manuela and Pirnog, Ionut},\n doi = {10.1109/EMES58375.2023.10171678},\n journal = {2023 17th International Conference on Engineering of Modern Electric Systems, EMES 2023}\n}","author_short":["Oprea, C., C.","Dragulinescu, A., M., C.","Marcu, I., M.","Pirnog, I."],"urls":{"Paper":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c/file/19093a12-9cea-08ab-43bb-d8f568f0f942/full_text.pdf.pdf"},"biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","bibbaseid":"oprea-dragulinescu-marcu-pirnog-evaluatingcriticaldiseaseoccurrenceingrapevineleavesusingcnnusecaseineasterneurope-2023","role":"author","keyword":["oprea2023evaluatingcriticaldisease"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/bfbbf840-4c42-3914-a463-19024f50b30c","dataSources":["2252seNhipfTmjEBQ"],"keywords":["oprea2023evaluatingcriticaldisease"],"search_terms":["evaluating","critical","disease","occurrence","grapevine","leaves","using","cnn","use","case","eastern","europe","oprea","dragulinescu","marcu","pirnog"],"title":"Evaluating Critical Disease Occurrence in Grapevine Leaves using CNN: Use-Case in Eastern Europe","year":2023}