Compressed Domain Watermarking via Bitstream Modification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing media encoding techniques require re-encoding of media streams for each user-specific watermark, making it impractical to add new users, as watermarks need to be applied before or during encoding in the picture domain, which is not feasible in practice.
Innovation Solution
A method for adding user-specific watermarks or data in the compressed domain using lightweight microstreams that modify the compressed bitstream, allowing for unique per-user bitstreams to be generated without re-encoding, using syntax elements and entropy coding to produce a decodable bitstream that can be imperceptibly modified or scrambled.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If watermarking is applied in the picture domain before or during encoding, then watermarking functionality is achieved, but re-encoding is required for each new user which is not feasible in practice
Solution Approach 1:
The patent separates the watermarking process from the encoding process by working in the compressed domain. Instead of modifying the picture domain data before encoding, the system operates on already-encoded bitstreams, segmenting the watermarking operation as a distinct post-processing step that does not require re-encoding the entire media stream.
Solution Approach 2:
The patent introduces an intermediary approach by using compressed domain representations as the medium for watermarking. Rather than directly modifying the original picture data or the final decoded output, the system embeds watermarks in the intermediate compressed bitstream format, allowing efficient user-specific customization without full re-encoding.
2Loss of information
If per-user watermarks are applied to original media, then unique identification for each user is achieved, but computational cost and time increase significantly
Solution Approach 1:
The patent applies preliminary action by encoding the media stream once in advance to produce the base compressed bitstream. User-specific watermarks are then added efficiently to this pre-encoded stream without requiring re-encoding, thus preserving user identification capability while minimizing processing time for new user additions.
Solution Approach 2:
The system creates lightweight copies of the compressed bitstream for each user with minimal modifications. Instead of re-processing the original media, the patent generates user-specific versions by copying and subtly modifying the already-encoded bitstream, dramatically reducing computational overhead while maintaining unique per-user identification.
3Adaptability or versatility
If media streams are re-encoded for each user, then user-specific customization is achieved, but device complexity and computational resources increase
Solution Approach 1:
The patent inverts the traditional approach by working backwards from the encoded bitstream rather than forwards from the original media. Instead of encoding customized content for each user, the system starts with a single encoded stream and applies minimal user-specific modifications, thereby achieving customization while dramatically reducing device complexity.
Solution Approach 2:
The patent achieves user-specific customization by changing parameters in the compressed domain rather than re-encoding the entire stream. By modifying specific bits or parameters in the already-encoded bitstream according to user identifiers, the system provides adaptability without the computational burden of full re-encoding operations.
Data Source
AI summary
A computer-implemented method is provided for processing a video stream in the compressed domain for watermarking, scrambling and other applications. Syntax elements are generated for input video as part of a video compression process. The syntax elements are entropy coded with an arithmetic entropy encoding process to produce a compressed bitstream for the input video. Regions of frames and related syntax elements of the input video are identified as candidates for modification. Based on metadata associated with a particular user, the syntax elements, the regions, and entropy coding state of the arithmetic entropy encoding process, bytes of the input video are changed to generate a modifying bitstream that is unique to the particular user; and modifying the compressed bitstream using the modifying bitstream to produce a decodable bitstream for the input video.