ocelli: Camera-Based TRNG

Ocelli is a Rust application that uses a camera to generate high-quality random entropy by analyzing pixel intensity differences between consecutive frames.

Features

• Camera-Based Entropy: Leverages grayscale pixel intensity differences to produce entropy bits.
• Shannon Entropy Validation: Filters entropy data to ensure high randomness quality.
• Van Neumann Whitening: Optional bias removal for unbiased entropy output.
• Dynamic Frame Handling: Automatically captures frames based on resolution and entropy requirements.

Requirements

• OpenCV (with Rust bindings)

Usage

Run the program with the desired parameters:

cargo run -- <entropy_length> <resolution_width> <resolution_height> [-w]

Parameters:

• <entropy_length>: Number of bytes of entropy to generate.
• <resolution_width>: Camera resolution width.
• <resolution_height>: Camera resolution height.
• -w: (Optional) Enable Van Neumann whitening for unbiased entropy.

Example:

Generate 1000 bytes of entropy at 1920x1080 resolution with whitening enabled:

cargo run -- 1000 1920 1080 -w

---

Output

• Hexadecimal Entropy: Generated entropy is output as a hexadecimal string.
• Processing Time: Reports how long the entropy generation took.
• Shannon Entropy: Displays the Shannon entropy of the final output.