🛰️ OrbitGuardian

Extreme-Performance Visual Aircraft Tracking & Trajectory Intelligence

37+ FPS / 4K Ultra-HD / Radar-Free Monitoring Production-grade aerospace computer vision that turns monocular video into real-time tactical intelligence.

---

💎 Key Features

• 🚀 4K Turbo Pipeline: Threaded I/O with FastVideoReader and FastVideoWriter achieving 37+ FPS average and 70+ FPS peak internal throughput.
• 👓 Crystal Clear HUD: UI rendered at native 3840x2160 (4K) resolution with resolution-aware dynamic scaling for precision and legibility.
• 🛡️ Digital Stabilization: Real-time image stabilization (DIS) to smooth out camera jitter and violent maneuvers.
• 🏎️ Flight Dynamics: Intelligent state detection including "STABLE", "SHARP TURN", "CLIMBING", and "EVASIVE" flight profiles.
• 🔮 Trajectory Prediction: LSTM-powered future path estimation with 15-step look-ahead accuracy.
• 📟 Tactical HUD: High-contrast aerospace overlay with speed (KMH/Mach), heading, and mission status telemetry.

---

🏗️ System Architecture

graph TD
    subgraph "I/O Layer (Threaded)"
        A[FastVideoReader]
        G[FastVideoWriter]
    end

    subgraph "Vision Engine"
        B[YOLOv8 Detection]
        C[Kalman Multi-Tracker]
        H[Digital Stabilization]
    end

    subgraph "Intelligence Layer"
        D[Dynamics Analyzer]
        E[LSTM Path Predictor]
        I[Speed/Motion Estimator]
    end

    subgraph "Presentation"
        F[4K Native HUD Engine]
    end

    A --> H
    H --> B
    B --> C
    C --> I
    I --> D
    D --> E
    E --> F
    F --> G

Loading

---

📁 Project Structure

orbitguardian/
├── analytics/            # Trajectory, Speed, & Dynamics Analysis
├── api/                  # FastAPI REST server for video processing
├── models/               # LSTM Prediction architecture & weights
├── tracking/             # Kalman filter & Multi-object tracker
├── vision/               # Detection Engine & Resolution-aware HUD
├── utils/                # Config, Auto-Device, & Central Logging
├── ui/                   # Minimalist HTML/JS Frontend Dashboard
├── outputs/              # 4K Processed videos & Flight Logs (CSV/JSON)
├── uploads/              # Raw data ingestion directory
└── main.py               # Core High-Performance Pipeline

---

🚀 Installation & Usage

⚙️ Setup

# Clone and setup environment
git clone https://github.com/darknecrocities/OrbitGuardian.git
cd orbitguardian
python -m venv venv
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

🏎️ Run Turbo 4K Pipeline

# Process video with real-time preview
python main.py --input samplevideo.mp4 --preview

# Run with custom output path
python main.py --input footage.mov --output result_4k.mp4

---

📈 Benchmarks (Phase 15 Tuning)

Metric	Target	Result (Turbo)
Average FPS	25.0	37.5 FPS
Peak Throughput	40.0	71.8 FPS
Output Resolution	1080p	2160p (4K)
Text Legibility	High	Crystal Clear (Native 4K)
Stability	Moderate	High (DIS Implementation)

---

🏗️ Development Roadmap (Completed)

1. Project Foundation & Structure
2. YOLOv8 Aircraft Detection
3. Kalman/Hungarian Tracking Engine
4. Trajectory Accumulation
5. Calibrated Speed Estimation
6. LSTM Trajectory Prediction
7. Aerospace HUD Overlay
8. Video Ingestion Pipeline
9. FastAPI REST Interface
10. Web Dashboard UI
11. Ego-motion & Radar Systems
12. DIS Stabilization & Legibility
13. Flight Dynamics Analysis
14. 4K Dynamic Scaling
15. Threaded Turbo Optimization

---

📊 Output Data

The system generates a high-fidelity CSV and JSON log for every flight containing:

• track_id, speed_kmh, heading_deg
• x, y (Detected) and future_x, future_y (Predicted)
• timestamp (Frame-accurate)

---

📄 License

MIT License. Created by Arron Kian Parejas.