/Potato-Leaf-Disease

Potato Leaf Disease Classification - Training using CNN.

Primary LanguageJupyter NotebookMIT LicenseMIT

Potato-Leaf-Disease Predictor

Problem Statement:-

"Every year, farmers face substantial economic losses and crop wastage due to diseases that affect potato plants. Two of the most devastating diseases that afflict potato leaves are Early Blight and Late Blight. These diseases are responsible for significant reductions in potato yields and can lead to severe economic hardship for farmers"

Thus the images are classified into 3 classes :

  • Healthy leaves

image

  • Late Blight leaves

image

  • Early Blight leaves

image

Technology stack used:-

  • Tensorflow
  • Numpy
  • Keras
  • Matplotlib
  • Pathlib

Dataset :-

The dataset is being taken from "Kaggle".

The dataset contains 15 different directories:-

  • Potato___Early_blight

  • Potato___Late_blight

  • Potato___healthy

each containing images in .jpg format.

Workflow:-

image

Output:-

image

The model accuracy after training the image dataset is > 97%.