HeliosIA

HeliosIA is an AI-powered tool designed to analyze nighttime road usage by cars and pedestrians using drone-captured videos. By processing high-altitude, downward-looking footage, HeliosIA maps traffic patterns, visualizes pedestrian movement, and provides valuable insights into road utilization under low-light conditions.

Table of Contents

• Features
• Prerequisites
• Installation
• Usage
• Processing Raw Videos
• Contributing
• License
• Acknowledgments

Features

• Traffic Analysis: Detects and tracks vehicles and pedestrians in nighttime drone footage.
• Visualization: Maps and visualizes traffic patterns and pedestrian movements.
• Insights: Provides data-driven insights on road utilization during low-light conditions.
• AI-Powered: Utilizes advanced AI algorithms for object detection and tracking.

Prerequisites

• Python 3.7 or higher
• Required Python packages (see requirements.txt)
• Processed video data (see Processing Raw Videos)

Installation

1. Clone the Repository

   git clone https://github.com/Septimus4/HeliosIA.git

2. Navigate to the Project Directory

   cd HeliosIA

3. Install Required Packages

   pip install -r requirements.txt

Usage

1. Prepare Processed Videos

   Ensure your drone-captured videos are processed using HeliosPostProcessing before analysis.

2. Run the Analysis

   python analyze.py --input path_to_processed_videos --output results/

3. View Results

   Analysis results will be saved in the specified output directory. Use provided visualization tools to interpret the data.

Processing Raw Videos

To process raw drone videos for use with HeliosIA, utilize the HeliosPostProcessing tool. This pre-processing step includes stabilizing footage, enhancing low-light visibility, and formatting data for analysis.

Steps:

1. Clone HeliosPostProcessing

   git clone https://github.com/Septimus4/HeliosPostProcessing.git

2. Process Videos

   Follow the instructions in the HeliosPostProcessing repository to prepare your videos.

3. Output

   Use the processed videos as input for HeliosIA.

Contributing

Contributions are welcome! Please open an issue or submit a pull request for enhancements or bug fixes.

License

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