Hydroponic Lettuce Disease Classification

training and evaluating a deep learning model to classify diseases in hydroponic lettuce.

Architecture

• Model: MobileNetV2 (Transfer Learning)
• Framework: TensorFlow/Keras
• Interface: Gradio Web App

Setup Instructions

1. Install Dependencies:

   pip install -r requirements.txt

2. Dataset Preparation: Place your dataset in a data/raw/ directory. The structure should be:

   data/
     raw/
       healthy/
         img1.jpg
         ...
       downy_mildew/
         ...
       powdery_mildew/
         ...
   

3. Training the Model:

   python -m src.train

   This will save the best model to models/best_model.keras and training plots to the outputs/ directory.

4. Evaluating the Model:

   python -m src.evaluate

   This generates a classification report and confusion matrix in the outputs/ directory.

5. Running the Web Interface:

   python -m src.app

   This starts a local web server to test the model on new images with confidence scores.