Grape Leaves Diseases Classification using Ensemble Learning and Transfer Learning

Abstract

Agriculture remains an important sector of the economy. Plant diseases and pests have a big impact on plant yield and quality. So, prevention and early detection of crop disease are some of the measures that must be implemented in farming to save the plants at an early stage and thereby reduce the overall food loss. Grapes are the most profitable fruit, but they are also vulnerable to a variety of diseases. Black Measles, Black Rot, and Leaf Blight are diseases that affect grape plants. Manual disease diagnosis can result in improper identification and use of pesticides, and it takes a long time. A variety of deep learning approaches have been used to address this issue of the identification and classification of grape leaf diseases. However, there are also limits to such approaches. Therefore, this paper uses deep learning with the concept of ensemble learning based on three famous Convolutional Neural Network (CNN) architectures (Visual Geometry Group (VGG16), VGG19, and Extreme Inception (Xception)). These three models are pretrained with ImageNet. The performance of the proposed approach is analyzed using the Plant Village (PV) dataset of common grape leaf diseases. The Proposed model gives higher performance than the results achieved by using each Deep Learning architecture separately and compared with the recent approaches in this study. The proposed system outperformed the others with 99.82% accuracy.

Classification of Plant Leaf Disease Using Ensemble Deep Transfer Learning

“…According to experiments, it was found that the performance of the proposed model in the second experiment, provided better accuracy with 99.38% and a loss of 0.0211 for training and an accuracy of 99.35 % and a loss of 0.019 for testing. This is likely due to the fact that the second experiment had fewer training sets than the first experiment, which allowed the model to learn more effectively and generalize better to new data, which goes beyond the testing accuracy of the studies done by (Tiwari, Divyansh, et al, [14]), (S. L. Sanga et al [17]), and (Hema et al [13] ) as shown in Table 5 With an accuracy of 99.82%, the proposed model's accuracy is just somewhat lower than that of the work of Andrew, et al [16]. However, only 37 million training parameters are utilized in the proposed work, compared to the number of training parameters used in [16].…”

“…This is likely due to the fact that the second experiment had fewer training sets than the first experiment, which allowed the model to learn more effectively and generalize better to new data, which goes beyond the testing accuracy of the studies done by (Tiwari, Divyansh, et al, [14]), (S. L. Sanga et al [17]), and (Hema et al [13] ) as shown in Table 5 With an accuracy of 99.82%, the proposed model's accuracy is just somewhat lower than that of the work of Andrew, et al [16]. However, only 37 million training parameters are utilized in the proposed work, compared to the number of training parameters used in [16]. This means that the proposed model requires fewer computational resources to train, can be trained faster, and reduce overfitting.…”

“…Nader, Andrew [16] This work combines the concept of ensemble learning with deep learning using three popular convolutional neural network models (VGG16, VGG19, and Xception). Grape leaf diseases is used to evaluate the suggested method.…”

Analisis Performa Inceptionv3 Convolutional Network Pada Klasifikasi Varietas Daun Grapevine

“…Pendekatan pembelajaran mendalam menggunakan pembelajaran ansambel berdasarkan tiga arsitektur Convolutional Neural Network (CNN) yang terkenal (Visual Geometry Group (VGG16), VGG19, dan Extreme Inception (Xception)) untuk klasifikasi penyakit daun anggur. Model yang diusulkan dilatih sebelumnya dengan ImageNet dan dianalisis menggunakan kumpulan data Plant Village (PV) dari penyakit daun anggur yang umum (Nader et al, 2022). Penelitian tentang klasifikasi varietas daun anggur menggunakan model DenseNet-30 menunjukkan hasil akurasi terbaik yaitu 98%, didalam penelitian ini akan menggunakan dataset yang sama dengan penelitian sebelumnya akan tetapi menggunakan metode yang berbeda, yaitu InceptionV3 diharapkan mencapai akurasi yang lebih baik.…”