Automated deep feature fusion based approach for the classification of multiclass rice diseases
Küçük Resim Yok
Tarih
2024
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Springer International Publishing
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Rice is a widely used grain product all over the world, and obtaining high-quality rice products is crucial. However, the quantity and quality of rice products can be reduced due to rice diseases. Detecting these diseases is challenging, as rice is cultivated in large, wet areas. Therefore, computer-aided systems for identifying rice diseases are of great significance. In this study, we propose a novel approach for detecting diseases in rice plants. Our approach employs three different convolutional neural networks (CNNs), namely Efficientb0, Shufflenet, and Resnet101. We extract feature maps from these networks, combine them, and then classify them using support vector machine (SVM). Additionally, seven different CNN architectures are employed to compare results. Our proposed approach achieves the highest accuracy value of 98%, demonstrating its potential for accurately classifying diseases in rice plants. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2023.
Açıklama
Anahtar Kelimeler
Artificial Intelligence, Cnn, Deep Learning, Rice Diseases, Svm
Kaynak
Iran Journal of Computer Science
WoS Q Değeri
Scopus Q Değeri
Q2
Cilt
7
Sayı
1
