A customizable video feed overlay system that creates a visual interface on camera feeds. It features animated grid cells that highlight when motion is detected, head tracking with animated reticles, and "informational displays" for identified persons.

After seeing the movie "Morgan", I wanted to make something neat like their camera overlay. This is just for fun.

From the movie "Morgan":

This system:

• Dynamic grid overlay with customizable dimensions
• Motion detection with animated grid cell highlighting
• Head tracking with animated reticle
• Person identification display panel
• Support for external motion detection data sources (ONVIF, object detection models)
• Highly customizable visual elements
• GPU acceleration support

# Using uv (recommended)
uv pip install -r requirements.txt

# Or using pip
pip install -r requirements.txt

# Run with default settings
./overlay_app.py

# Run with custom settings
./overlay_app.py --source 0 --grid-cols 9 --grid-rows 5

# Run with video example
./overlay_app.py --source ./path/to/video.mp4

Configuration can be done via command line arguments or by editing the config.yaml file.

See config.yaml for all available options.

• AMD GPUs (via OpenCL)
• NVIDIA GPUs (via CUDA)
• NVIDIA Jetson (optimized support)

Docker configurations are available for different hardware platforms:

• docker-compose.yml - Basic setup
• docker-compose.nvidia.yml - NVIDIA GPU support
• docker-compose.jetson.yml - Jetson device support