Multi Stage Neural Network Based Ensemble Learning Approach for Wheat Leaf Disease Classification
Multi Stage Neural Network Based Ensemble Learning Approach for Wheat Leaf Disease Classification
Mamanduru Sunil
M-Tech, Department .Of Computer Scince And
Engineering,
Vemu Institute Of Technology,
P.Kothakota,Chittoor District, Andhra Pradesh
517112,India
Email Id: Mamandurusunil2000@Gmail.Com
Mr.Dharmaiahvari prasad
Assistant professor, M.Tech,Dept of CSE,
Vemu institute of Technology ,p.kothakota.
Email Id: dprasadnaresh@gmail.com
Abstract: The proposed automated system in the paper relies on the notion of deep learning and the proposed functionality is to identify the diseases of the wheat plants with the help of Wheat Plant Diseases dataset that could be downloaded without cost on Kaggle. The paperproves the effectiveness of the test and the analysis of the work of one of the most successful representatives of the convolutional neural network (CNN) the ResNet50, DenseNet201, VGG19, NASNetLarge, Swin Transformer, and Vision Transformers (ViT). These models they are trained and fitted on are used to classify the diseases of wheat plants Aphid, Black Rust, Blast, Fusarium Head Blight, Yellow Rust etc. The evaluation of such models is also done in terms of accuracy and precision of models, recall and F1-score of models. The obtained results are evidence that DenseNet201 and Swin Transformer were the only networks that achieved the highest accuracy of 93 and the other networks possessed the following determinacy (ResNet50 (75), VGG19 (91), ViT (90), NASNetLarge (88) and CNN (90)). The paper shall make commentary on the future of deep learning as applied in precision agriculture which is a reliable and feasible mechanism in the detection of diseases in wheat plants at the appropriate time. Not only will this go a long way in helping the farmers and other agricultural professionals to reduce the losses of crops but also to greater awareness on yield and further crop protection.
Keywords: Deep Learning, CNN, ResNet50,DenseNet201, VGG19, Nasnetlarge, MobileNetV2,Vision transformers, Swin transformer, Plant pathology,Image classification, Precision agriculture, agriculturalAI, Transfer learning.