Crop Leaf Disease Identification Using Deep Transfer Learning
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Traditional manual identification of crop leaf diseases is challenging. Owing to the limitations in manpowerand resources, it is challenging to explore crop diseases on a large scale. The emergence of artificial intelligencetechnologies, particularly the extensive application of deep learning technologies, is expected to overcome thesechallenges and greatly improve the accuracy and efficiency of crop disease identification. Crop leaf diseaseidentification models have been designed and trained using large-scale training data, enabling them to predictdifferent categories of diseases from unlabeled crop leaves. However, these models, which possess strongfeature representation capabilities, require substantial training data, and there is often a shortage of such datasetsin practical farming scenarios. To address this issue and improve the feature learning abilities of models, thisstudy proposes a deep transfer learning adaptation strategy. The novel proposed method aims to transfer theweights and parameters from pre-trained models in similar large-scale training datasets, such as ImageNet. ImageNet pre-trained weights are adopted and fine-tuned with the features of crop leaf diseases to improveprediction ability. In this study, we collected 16,060 crop leaf disease images, spanning 12 categories, fortraining. The experimental results demonstrate that an impressive accuracy of 98% is achieved using theproposed method on the transferred ResNet-50 model, thereby confirming the effectiveness of our transferlearning approach.
Abstract
Traditional manual identification of crop leaf diseases is challenging. Owing to the limitations in manpowerand resources, it is challenging to explore crop diseases on a large scale. The emergence of artificial intelligencetechnologies, particularly the extensive application of deep learning technologies, is expected to overcome thesechallenges and greatly improve the accuracy and efficiency of crop disease identification. Crop leaf diseaseidentification models have been designed and trained using large-scale training data, enabling them to predictdifferent categories of diseases from unlabeled crop leaves. However, these models, which possess strongfeature representation capabilities, require substantial training data, and there is often a shortage of such datasetsin practical farming scenarios. To address this issue and improve the feature learning abilities of models, thisstudy proposes a deep transfer learning adaptation strategy. The novel proposed method aims to transfer theweights and parameters from pre-trained models in similar large-scale training datasets, such as ImageNet. ImageNet pre-trained weights are adopted and fine-tuned with the features of crop leaf diseases to improveprediction ability. In this study, we collected 16,060 crop leaf disease images, spanning 12 categories, fortraining. The experimental results demonstrate that an impressive accuracy of 98% is achieved using theproposed method on the transferred ResNet-50 model, thereby confirming the effectiveness of our transferlearning approach.
참고문헌 (18)
- N. R. Bhimte, "Diseases detection of cotton leaf spot using image processing and SVM classifier", Proceedings of 2018 2nd International Conference on Intelligent Computing and Control Systems (ICICCS), pp.340~344 (2018)
- D. A. Padilla, "Portable yellow spot disease identifier on sugarcane leaf via image processing using support vector machine", Proceedings of 2019 5th International Conference on Control, Automation and Robotics (ICCAR), pp.901~905 (2019)
- P. Sharma, "Classification of plant leaf diseases using machine learning and image preprocessing techniques", Proceedings of 2020 10th International Conference on Cloud Computing, Data Science & Engineering (Confluence), pp.480~484 (2020)
- C. Zhou, "COVID-19 detection based on image regrouping and ResNet-SVM using chest X-ray images", IEEE Access, 9, pp.81902~81912 (2021)[DOI]
- M. Z. Hasan, "Recognition of jute diseases by leaf image classification using convolutional neural network", Proceedings of 2019 10th International Conference on Computing, Communication and Networking Technologies (ICCCNT), pp.1~5 (2019)
- J. G. Barbedo, "Factors influencing the use of deep learning for plant disease recognition", Biosystems Engineering, 172, pp.84~91 (2018)[DOI]
- S. Ashok, "Tomato leaf disease detection using deep learning techniques", Proceedings of 2020 5th International Conference on Communication and Electronics Systems (ICCES), pp.979~983 (2020)
- C. Zhou, "Tomato leaf disease identification by restructured deep residual dense network", IEEE Access, 9, pp.28822~28831 (2021)[DOI]
- Q. H. Cap, "LeafGAN : an effective data augmentation method for practical plant disease diagnosis", IEEE Transactions on Automation Science and Engineering, 19(2), pp.1258~1267 (2022)[DOI]
- P. Zhao, "Robust transfer learning based on geometric mean metric learning", Knowledge-Based Systems, 227 (2021)[DOI]
- S. Coulibaly, "Deep neural networks with transfer learning in millet crop images", Computers in Industry, 108, pp.115~120 (2019)[DOI]
- A. Abd Almisreb, "Utilizing AlexNet deep transfer learning for ear recognition", Proceedings of 2018 4th International Conference on Information Retrieval and Knowledge Management (CAMP), pp.1~5 (2018)
- X. Zhang, "Accelerating very deep convolutional networks for classification and detection", IEEE Transactions on Pattern Analysis and Machine Intelligence, 38(10), pp.1943~1955 (2016)[DOI]
- T. Emara, "A modified Inception-v4 for imbalanced skin cancer classification dataset", Proceedings of 2019 14th International Conference on Computer Engineering and Systems (ICCES), pp.28~33 (2019)
- K. He, "Deep residual learning for image recognition", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp.770~778 (2016)
- G. Huang, "Densely connected convolutional networks", Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp.2261~2269 (2017)
- "ImageNet large scale visual recognition challenge 2012 (ILSVRC2012)" (2020)[URL]
- J. Long, "Fully convolutional networks for semantic segmentation", IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(4), pp.640~651 (2017)[DOI]