This project is a Potato Disease Classification Web App built using TensorFlow/Keras and deployed with Streamlit.
The model can classify potato leaf images into three categories:
- Healthy
- Early Blight
- Late Blight
The app allows users to upload an image or take a picture using their camera, and the system will predict the disease along with the model confidence score.
The dataset used for training comes from the PlantVillage Dataset on Kaggle.
For this project, only the following subsets were used:
Potato___healthyPotato___Early_blightPotato___Late_blight
This dataset contains thousands of high-quality, labeled plant leaf images widely used for agricultural disease detection research.
The model was built using TensorFlow Keras with the following setup:
- Architecture: Convolutional Neural Network (CNN)
- Optimizer: Adam (chosen for stable and fast convergence)
- Loss Function: Categorical Crossentropy
- Epochs: 15
- Final Training Performance:
- Accuracy:
0.9730 - Loss:
0.0514
- Accuracy:
These results indicate that the model is well-optimized and performs strongly on the classification task.
The web interface is built with Streamlit, providing a simple and clean user experience.
Live Demo : https://styfie-potato-disease-classification-app-f88a8h.streamlit.app/
- Upload a potato leaf image
- Take a picture using webcam
- Model predicts whether the leaf is:
- Healthy
- Early Blight
- Late Blight
- Displays prediction confidence
- Responsive, minimalistic UI
-
Clone the repository:
git clone https://github.com/your-username/repo-name.git
-
Install dependencies:
pip install -r requirements.txt
-
Run the Streamlit app:
streamlit run app.py
π¦ Potato Disease Classification
β 1.keras
β£ app.py
β£ requirements.txt
β README.md
This project is for educational and research purposes.