🛡️ Smart Barrier AI

](https://github.com/ituvtu/smart-barrier-ai)

Smart Barrier AI is an intelligent vehicle access control system designed to automate gate operations. It utilizes a fine-tuned YOLOv10m model for license plate detection and EasyOCR for text recognition, enhanced by robust image preprocessing and region-specific post-processing logic.

🚀 Key Features

• High-Precision Detection: Uses a custom-trained YOLOv10m model to accurately locate license plates in various lighting conditions.
• Advanced OCR Pipeline: Implements EasyOCR with a custom preprocessing stage (upscaling, denoising, adaptive thresholding) to handle low-quality images.
• Region Enforcer Logic:
  • 🇺🇦 Ukraine: Enforces specific standard structure (LL DDDD LL) with sliding window analysis and character correction.
  • 🇪🇺 Europe: Adaptive filtering for EU plates (stripping country codes and artifacts).
• Fuzzy Matching: Smart database verification that accounts for common OCR typos (e.g., 0 vs O, 8 vs B).
• Interactive UI: A user-friendly web interface built with Gradio featuring simulated LED status indicators and real-time logs.

🛠️ Installation

1. Clone the repository:

   git clone https://github.com/ituvtu/smart-barrier-ai.git
   cd smart-barrier-ai

2. Install dependencies:

   pip install -r requirements.txt

3. Verify Model Weights: Ensure that the fine-tuned model weights file best.pt is located in the root directory.

💻 Usage

Run the application locally:

python app.py

Open your browser and navigate to the local URL provided in the terminal (usually http://127.0.0.1:7860).

Workflow:

1. Upload an image of a vehicle.
2. Select the region mode (e.g., Ukraine 🇺🇦).
3. Populate the "Database" field with allowed license plates (comma-separated).
4. Click "Check Access" to see the detection result and access decision.

📂 Project Structure

• app.py - Main entry point containing the Gradio interface and pipeline orchestration.
• utils.py - Core logic for text cleaning, regex pattern matching, and fuzzy search algorithms.
• requirements.txt - Python dependencies.
• best.pt - Fine-tuned YOLOv10m weights (ensure this file is present).

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Author: ituvtu