Video Frame Watermarking for Tamper Detection Over Lossy Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video processing technologies, particularly with advancements in deep learning, make it difficult to distinguish between original and tampered video sequences, especially over lossy channels, necessitating a secure method to authenticate video frames.
Innovation Solution
Embedding an encrypted signature or watermark in video frames using low frequency metrics resistant to lossy compression, which can be decrypted for authentication, ensuring the integrity of the video content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If video is transmitted over lossy channels using compression techniques, then transmission efficiency is improved, but video authenticity and integrity are compromised
Solution Approach 1:
The patent embeds an encrypted signature or watermark into the video frame during the encoding process, before transmission. This preliminary action ensures that authentication data is already present in the transmitted video, allowing verification without requiring separate authentication channels or post-transmission processing.
Solution Approach 2:
The patent uses an encrypted signature or watermark as an intermediary element that carries authentication information within the video frame itself. This intermediary allows the video to maintain both compression efficiency and authenticity verification capability by separating the authentication function from the video content while embedding both in the same data stream.
2Reliability
If watermarks are embedded in video frames, then authentication capability is improved, but image quality and compression performance deteriorate
Solution Approach 1:
The patent embeds the watermark by modifying specific pixel values at particular locations within the video frame, rather than uniformly affecting the entire image. This localized modification approach allows the watermark to be embedded with minimal impact on overall image quality, as only specific regions are altered to encode the authentication data.
Solution Approach 2:
The patent embeds the watermark by subtly changing pixel value parameters within the video frame, using encryption to distribute the watermark data across multiple pixels. This parameter change approach allows the watermark to be embedded in a way that is resistant to compression while maintaining visual quality, as the modifications are distributed and encrypted rather than concentrated.
3Reliability
If existing watermarking techniques are used, then authentication is possible, but the methods are vulnerable to deep learning-based tampering and lossy compression
Solution Approach 1:
The patent applies encryption to the signature or watermark during the encoding process, before transmission. This preliminary encryption action ensures that the authentication data is protected against tampering and cannot be easily manipulated by deep learning-based attacks, as the encrypted nature of the watermark makes it resistant to modification while remaining robust to lossy compression.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques are provided for generation of secure video and tamper detection of the secure video. A methodology implementing the techniques according to an embodiment includes selecting a subset of macroblocks from a video frame to be transmitted and calculating a low frequency metric on each of the selected macroblocks. The method also includes performing a hash calculation on the low frequency metrics to generate a frame signature; encrypting the frame signature (using a private key) to generate an encrypted watermark; and modifying pixels of each of the selected macroblocks to generate the secured video frame, the modifications based on bits of the encrypted watermark that are associated with the selected macroblock. The method further includes authenticating a received video frame by comparing a calculated frame signature to an authenticated frame signature, the authenticated frame signature decrypted (using a public key) from an extracted watermark of the received video frame.