A real time monitoring system for accurate plant leaves disease detection using deep learning
Accurate and timely detection of plant diseases is crucial for sustainable agriculture and food security. This research presents a real-time monitoring system utilizing deep learning techniques to detect diseases in plant leaves with high accuracy. We combined several plant datasets, including the P...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-02-01
|
| Series: | Crop Design |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2772899424000417 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589917938515968 |
|---|---|
| author | Kazi Naimur Rahman Sajal Chandra Banik Raihan Islam Arafath Al Fahim |
| author_facet | Kazi Naimur Rahman Sajal Chandra Banik Raihan Islam Arafath Al Fahim |
| author_sort | Kazi Naimur Rahman |
| collection | DOAJ |
| description | Accurate and timely detection of plant diseases is crucial for sustainable agriculture and food security. This research presents a real-time monitoring system utilizing deep learning techniques to detect diseases in plant leaves with high accuracy. We combined several plant datasets, including the PlantVillage Dataset, resulting in a comprehensive dataset of 30,945 images across eight plant types (potato, tomato, pepper bell, apple, corn, grape, peach, and rice) and 35 disease classes. Initially, a custom Convolutional Neural Network (CNN) model was developed, achieving a leaf classification accuracy of 95.62 %. Subsequently, the dataset was partitioned for individual plant disease detection, applying nine different CNN models (custom CNN, VGG16, VGG19, InceptionV3, MobileNet, DenseNet121, Xception, and two hybrid models) to each plant type. The highest accuracy rates for disease detection were: 100 % for potato (custom CNN), 98 % for tomato (InceptionV3, custom CNN, VGG16), 100 % for pepper bell (MobileNet, custom CNN), 100 % for apple (MobileNet, Xception), 98 % for corn (custom CNN), 99 % for grape (custom CNN, VGG19, DenseNet121), 100 % for peach (VGG16, custom CNN), and 98 % for rice (DenseNet121). A web and mobile application were developed based on the best-performing models, allowing users to insert or capture images of plant leaves, detect diseases, and receive treatment suggestions with high confidence levels. The results demonstrate the effectiveness of deep learning models in accurately identifying plant diseases, offering a valuable tool for enhancing disease management and crop yields. |
| format | Article |
| id | doaj-art-7b70400ffb9b499fbdd1573a6c5906f8 |
| institution | Kabale University |
| issn | 2772-8994 |
| language | English |
| publishDate | 2025-02-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Crop Design |
| spelling | doaj-art-7b70400ffb9b499fbdd1573a6c5906f82025-01-24T04:46:04ZengElsevierCrop Design2772-89942025-02-0141100092A real time monitoring system for accurate plant leaves disease detection using deep learningKazi Naimur Rahman0Sajal Chandra Banik1Raihan Islam2Arafath Al Fahim3Corresponding author.; Chittagong University of Engineering & Technology, CUET, 4349, Chittagong, BangladeshChittagong University of Engineering & Technology, CUET, 4349, Chittagong, BangladeshChittagong University of Engineering & Technology, CUET, 4349, Chittagong, BangladeshChittagong University of Engineering & Technology, CUET, 4349, Chittagong, BangladeshAccurate and timely detection of plant diseases is crucial for sustainable agriculture and food security. This research presents a real-time monitoring system utilizing deep learning techniques to detect diseases in plant leaves with high accuracy. We combined several plant datasets, including the PlantVillage Dataset, resulting in a comprehensive dataset of 30,945 images across eight plant types (potato, tomato, pepper bell, apple, corn, grape, peach, and rice) and 35 disease classes. Initially, a custom Convolutional Neural Network (CNN) model was developed, achieving a leaf classification accuracy of 95.62 %. Subsequently, the dataset was partitioned for individual plant disease detection, applying nine different CNN models (custom CNN, VGG16, VGG19, InceptionV3, MobileNet, DenseNet121, Xception, and two hybrid models) to each plant type. The highest accuracy rates for disease detection were: 100 % for potato (custom CNN), 98 % for tomato (InceptionV3, custom CNN, VGG16), 100 % for pepper bell (MobileNet, custom CNN), 100 % for apple (MobileNet, Xception), 98 % for corn (custom CNN), 99 % for grape (custom CNN, VGG19, DenseNet121), 100 % for peach (VGG16, custom CNN), and 98 % for rice (DenseNet121). A web and mobile application were developed based on the best-performing models, allowing users to insert or capture images of plant leaves, detect diseases, and receive treatment suggestions with high confidence levels. The results demonstrate the effectiveness of deep learning models in accurately identifying plant diseases, offering a valuable tool for enhancing disease management and crop yields.http://www.sciencedirect.com/science/article/pii/S2772899424000417Plant disease detectionConvolutional neural networkDeep learning modelComputer visionPerformance evaluationImage processing |
| spellingShingle | Kazi Naimur Rahman Sajal Chandra Banik Raihan Islam Arafath Al Fahim A real time monitoring system for accurate plant leaves disease detection using deep learning Crop Design Plant disease detection Convolutional neural network Deep learning model Computer vision Performance evaluation Image processing |
| title | A real time monitoring system for accurate plant leaves disease detection using deep learning |
| title_full | A real time monitoring system for accurate plant leaves disease detection using deep learning |
| title_fullStr | A real time monitoring system for accurate plant leaves disease detection using deep learning |
| title_full_unstemmed | A real time monitoring system for accurate plant leaves disease detection using deep learning |
| title_short | A real time monitoring system for accurate plant leaves disease detection using deep learning |
| title_sort | real time monitoring system for accurate plant leaves disease detection using deep learning |
| topic | Plant disease detection Convolutional neural network Deep learning model Computer vision Performance evaluation Image processing |
| url | http://www.sciencedirect.com/science/article/pii/S2772899424000417 |
| work_keys_str_mv | AT kazinaimurrahman arealtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT sajalchandrabanik arealtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT raihanislam arealtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT arafathalfahim arealtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT kazinaimurrahman realtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT sajalchandrabanik realtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT raihanislam realtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning AT arafathalfahim realtimemonitoringsystemforaccurateplantleavesdiseasedetectionusingdeeplearning |