Computer vision has recently developed into a crucial tool in agriculture for tracking plant health and treating illnesses. In this paper, we present a VGG19-based deep learning technique for recognising diseases in Potato leaves. Bacterial spot, early blight, late blight, and spider mite damage are four prevalent potato leaf diseases that the suggested system can detect automatically. 1,000 photos per class make up the 4,000 images of Potato leaves in the dataset used for this study. The photos in the dataset were resized and divided into training, validation, and testing sets as part of the pre-processing step. On the Potato leaf dataset, we refined the pre-trained VGG19 model using the transfer learning technique. According to experimental data, our suggested strategy outperformed other cutting-edge techniques, achieving an overall accuracy of 95.2% on the test set. Additionally, we compared various deep learning architectures and discovered that VGG19 is the best architecture for this task. A significant tool for farmers and agronomists, the suggested Potato leaf disease detection system might help them identify illnesses early and take the necessary precautions to stop their spread.
Paper
Full text
A Novel Framework for Potato Leaf Disease Detection Using Deep Learning Model
Semantic Scholar · Agricultural and Food Sciences · 2023
Abstract
Computer vision has recently developed into a crucial tool in agriculture for tracking plant health and treating illnesses. In this paper, we present a VGG19-based deep learning technique for recognising diseases in Potato leaves. Bacterial spot, early blight, late blight, and spider mite damage are four prevalent potato leaf diseases that the suggested system can detect automatically. 1,000 photos per class make up the 4,000 images of Potato leaves in the dataset used for this study. The photos in the dataset were resized and divided into training, validation, and testing sets as part of the pre-processing step. On the Potato leaf dataset, we refined the pre-trained VGG19 model using the transfer learning technique. According to experimental data, our suggested strategy outperformed other cutting-edge techniques, achieving an overall accuracy of 95.2% on the test set. Additionally, we compared various deep learning architectures and discovered that VGG19 is the best architecture for this task. A significant tool for farmers and agronomists, the suggested Potato leaf disease detection system might help them identify illnesses early and take the necessary precautions to stop their spread.