<<<<<<< HEAD

🔥 Forest Fire Detection Using AI

This project detects forest fires using Convolutional Neural Networks (CNN) and performs real-time fire detection via a mobile camera. It can optionally send email alerts when fire is detected and includes geolocation information.

📁 Project Structure

Forest-Fire-Detection-AI/
├── model_training/
│   ├── Forest_Fire_Detection_Train_Test_model.ipynb      # CNN model training notebook (Google Colab)           
├── live_detection/
│   ├── fire_detection_live.py      # Real-time fire detection using webcam
├── requirements.txt                # Python package dependencies
├── README.md                       # Project overview and usage
└── .gitignore                      # Files to ignore in GitHub

🧠 Model Details

• Model: Custom CNN
• Input size: 224x224 pixels RGB
• Output: Binary classification (Fire / No Fire)

🏗️ How to Use

🔹 1. Train the Model (Colab)

Run Forest_Fire_Detection_Train_Test_model in Google Colab to train the model on fire and non-fire images. You can save the model as .keras.

📱 2. Using Mobile as a Webcam

This project also supports using a mobile phone as a webcam using the IP Webcam app (Android):

1. Install the IP Webcam app from the Play Store.
2. Connect your mobile and laptop to the same Wi-Fi or mobile hotspot.
3. Open the app, scroll down, and tap "Start Server".
4. Note the streaming URL (e.g., http://192.168.0.101:8080/video).
5. In fire_detection_live.py, replace the webcam line:

cap = cv2.VideoCapture("http://192.168.0.101:8080/video")


### 🔹 3. Real-Time Detection (VS Code or Local Python)
Use `nforest_fire_detection.py` to detect fire from your mobile camera.

```bash
python nforest_fire_detection.py

Press q to quit the webcam stream.

🔥 Download Pre-trained Model

To use the real-time fire detection without training, download the pre-trained model from the link below:

👉 Download fire_model.keras

Place the downloaded file in the project folder before running live_detection.py.

📦 Installation

pip install -r requirements.txt

📥 Dataset

• Forest Fire Dataset (Kaggle)
• You can also use your own dataset with two folders: Fire/ and Non-Fire/

✅ Requirements

• TensorFlow 2.x
• OpenCV
• NumPy
• Matplotlib

(Install via requirements.txt)

🔐 Notes

• Don't upload large datasets to GitHub. Use links instead.
• Store sensitive credentials (like email passwords) securely.

👨‍💻 Author

VENKAT RAMANAN R
B.Tech ECE, VIT-AP University
Edunet-Microsoft AI Internship Project

📃 License

This project is open-source under the MIT License.

Forest-Fire-Detection-AI

🔥 AI-based forest fire detection system using CNN and real-time webcam monitoring with email alerts

9d1a01909a67ae67aa827c7eb83bf5252d045441