Menasha123/gvisions

🌿 Plant Disease Detection API

A Flask-based API for predicting plant diseases from images or numerical features using a pre-trained deep learning model. It also supports storing geolocation data related to detected diseases in Firebase Firestore.

🚀 Features

• —Predict plant diseases from uploaded leaf images.
• —Predict diseases based on numeric features.
• —Store disease name and associated geolocations.
• —Save results to Firebase Firestore.
• —Supports basic error handling and logging.

🛠️ Setup

1. Clone the Repository

git clone https://github.com/your-username/plant-disease-api.git
cd plant-disease-api

2. Install Dependencies

pip install -r requirements.txt

Key Libraries:

• —Flask
• —TensorFlow
• —NumPy
• —Firebase Admin SDK

3. Prepare Your Files

• —Place your trained model in the root directory as model.h5.
• —Include your Firebase service account JSON as ./serviceAccountKey.json.

📦 API Endpoints

/

Method: GET Description: Returns the home page (index.html).

/predict-image

Method: POST Content-Type: multipart/form-data Description: Predicts disease from an uploaded image.

Form Data:

• —file: image file (jpg/png)

Response:

{
  "predicted_class": "Mosaic Virus Disease",
  "confidence": 0.97,
  "all_predictions": {
    "Healthy Leaf": 0.01,
    ...
  }
}

Note: After prediction, the disease name is stored in memory to be associated with geolocation data.

/geolocation

Method: POST Content-Type: application/json Description: Stores geolocation data (latitude & longitude). If a disease is already predicted, it saves both to Firestore.

Request Body:

{
  "latitude": 12.9716,
  "longitude": 77.5946
}

Response:

{
  "message": "Geolocation data stored successfully"
}

🧠 Internal Functions

predict_disease(img_data)

• —Preprocesses the image.
• —Predicts disease using the loaded model.
• —Returns the class label, confidence, and full class probabilities.

save_disease_locations()

• —Combines the disease and geolocation data.
• —Saves the information to Firestore in the structure:

{
  "Disease Name": {
    "coordinates": {
      "01": {"latitude": 12.34, "longitude": 56.78},
      "02": ...
    }
  }
}

🔥 Firebase Integration

• —Initialized using serviceAccountKey.json.
• —Stores data in disease_coordinates collection.
• —Automatically merges new geolocations with existing ones.

❗ Error Handling

• —404 Not Found
• —500 Server Error
• —Custom validation for input types

🧪 Running the App

python app.py

Visit: http://127.0.0.1:5000

🧼 Utility Functions

• —clean_label(label): Removes "Augmented" prefix.
• —preprocess_image(img_data): Resizes and preprocesses images using ResNet50 conventions.