/PotatoLeafDisease

This repository contains an implementation of a CNN model leveraging transfer learning with Xception architecture. The model classify potato leaves into Healthy, Early Blight, and Late Blight diseases.

Primary LanguageJupyter NotebookMIT LicenseMIT

Potato Leaf Diseases

This repository features a CNN model employing transfer learning with Xception architecture to classify potato leaves into Healthy, Early Blight, and Late Blight diseases.

The model is trained on a dataset sourced from Kaggle. The data was collected from Rashid et al. Multi-Level Deep Learning Model for Potato Leaf Disease Recognition. Electronics. 2021; 10(17):2064. DOI: https://doi.org/10.3390/electronics10172064

Dataset Details:

  • Early Blight: Images of potato leaves infected with Early Blight (Alternaria solani), a fungal disease characterized by small, dark lesions with concentric rings. Early detection of this disease is crucial to prevent yield losses.
  • Late Blight: Images of potato leaves infected with Late Blight (Phytophthora infestans), a devastating disease that causes dark, irregularly shaped lesions on leaves and stems. Late Blight can lead to complete crop loss if not managed promptly and effectively.
  • Healthy Leaves: Images of potato leaves that exhibit no symptoms of disease. Healthy leaves serve as a reference category for training machine learning models.

© 2024 Developed by Anne Livia.