Abstract:

This study investigates spatial domain video steganography, a technique that conceals information by modifying pixel color values in consecutive video frames. The effectiveness of this method depends on embedding these modifications within the natural noise present in videos, which arises from inter-frame differences, compression artifacts, and processing variations. To inform and enhance steganographic techniques, we analyze the statistical properties of naturally occurring pixel color differences across frames.

A diverse sample of video files was examined, focusing on histograms and frequency patterns of color difference variations. The analysis revealed distinct patterns in color changes, indicating predictable natural variations between frames. Understanding these patterns is essential for developing encoding methods that exploit inherent noise, thereby reducing the detectability of hidden data and preserving the integrity of the video content.

The study contributes both theoretically and practically: theoretically, by providing insights into the statistical behavior of pixel color differences; and practically, by enhancing and guiding the development of spatial domain video steganography techniques.