🌿 PlantDoc-AI

(Deep Learning + FastAPI + Streamlit)

An Plant Disease Detection and Treatment System using PyTorch, ResNet-18, FastAPI REST API, and Streamlit UI. Upload a leaf image and instantly receive:

✔ Disease name ✔ Disease description ✔ Chemical treatment recommendations ✔ Bio-organic treatment suggestions ✔ Confidence score

---

📺 Live Demo (GIF)

🌐 Live Streamlit App

🚀 Try the Plant Disease Detection App here: 👉 https://plant-pest-disease-detection.streamlit.app/

✨ Features

🤖 Deep Learning Model

• Fine-tuned ResNet-18 classifier
• Trained in Google Colab (dataset loading, augmentation, splits, training loops visible in your training PDF)
• Clean prediction formatting
• High-confidence softmax output

🚀 FastAPI Backend

• /predict/ REST endpoint

• Accepts image → returns disease info

• Handles:

  • disease formatting
  • treatment cleaning
  • bio-organic treatment splitting

🎨 Streamlit Frontend

• User uploads image
• Displays model output neatly
• Shows treatments as bullet lists
• Clean centered UI

📄 Treatment Dataset Integration

• Reads Treatment_dataset.csv

• Fetches:

  • disease description
  • chemical treatments
  • bio/organic solutions

---

📁 Project Structure

📦 plant-disease-detection
├── backend/
│   ├── main.py                 # FastAPI backend
│   ├── plant_disease_model.pth # Trained model
│   ├── Treatment_dataset.csv   # Treatment data
├── frontend/
│   ├── app.py                  # Streamlit UI
├── notebooks/
│   ├── plant_disease_training.pdf
├── assets/
│   ├── plant_disease_demo.gif  # (Add your GIF here)
└── README.md

---

🧠 Model Training Summary

Your included training notebook (PDF) contains:

📌 Dataset Loading

• Reads plant disease folders
• Maps classes automatically

📌 Visualization

• Displays sample leaves (page 3 of your notebook)

📌 Data Splits

• Train / Test / Validation (stratified)

📌 Transformations

• Resize
• Random Flip
• Random Rotation
• Normalization
• (Same transforms used in prediction)

📌 Model Architecture

• ResNet-18
• Custom final linear layer replacing the original
• Adam optimizer
• CrossEntropyLoss

📌 Training Loop

• Forward → backward → optimizer step
• Epoch logging for training/validation

---

🚀 Getting Started

1️⃣ Clone the Repository

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

2️⃣ Install Dependencies

pip install -r requirements.txt

Dependencies include:

fastapi
uvicorn
torch
torchvision
pandas
pillow
streamlit

---

🧪 Running the Backend (FastAPI)

cd backend
uvicorn main:app --host 0.0.0.0 --port 8000

API Endpoint Example

POST /predict/ Send an image → receive prediction.

Response:

{
  "prediction": {
    "disease": "Strawberry - Leaf Scorch",
    "description": "Fungal disease causing brown spots...",
    "treatments": [
      "Copper oxychloride 50% WP",
      "Thiophanate methyl 70% WP"
    ],
    "bio_treatments": [
      "Neem oil 5ml/liter",
      "Trichoderma viride 2g/liter"
    ]
  },
  "confidence": 0.96
}

---

🎨 Running the Frontend (Streamlit)

cd frontend
streamlit run app.py

Streamlit UI will open in the browser.

---

🛠 Tech Stack

Component	Technology
Model Training	PyTorch
Backend	FastAPI
Frontend	Streamlit
Deployment	Streamlit Cloud
Dataset	Custom Plant Disease Dataset + Treatment CSV

---