| Abstrakt: |
The Indian economy relies heavily on the agricultural sector. Improving crop and plant yields is crucial because 60% of India's labour force is involved in this industry. It took until recently for Indian farmers to see increases in both production and selling prices due to a variety of crop-related ailments. The current revolution in smartphone penetration and computer vision models has opened up new opportunities for agricultural picture classification. Modern picture identification systems, such as Convolutional Neural Networks, are able to make precise and rapid diagnoses. In order to correctly detect plant diseases, this article utilizes pre-trained models that are built on convolutional neural networks (CNNs). Tuned-hyper-parameters for ResNet50, DenseNetl21, VGG16, and Inception V4 in particular. The Plant-Village dataset, which includes several image examples of various plant diseases, was utilized in the experiments. Fl, sensitivity, specificity, and classification accuracy were the parameters used to estimate the model's efficacy. We also compared our results to those of other, similar, state-of-the-art investigations. We can see that DenseNet-121 gets a success rate of 99.81% from the validation data. This paves the way for artificial intelligence solutions for small holder farmers and shows that convolutional neural networks (CNNs) can classify plant illnesses. [ABSTRACT FROM AUTHOR] |
