Video Watermark Channel Decomposition for Geometric Robustness
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video watermarking technologies face inefficiencies in encoding and decoding, particularly for high-resolution videos, leading to increased computing power consumption and reduced image quality, and are vulnerable to geometric transformations that compromise decoding robustness.
Innovation Solution
Embed watermark information and location information into different channels of local image blocks within a video frame using channel decomposition, without altering the video resolution, to enhance encoding efficiency and robustness against geometric transformations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If watermark information is embedded into the entire target frame image, then watermark coverage is improved, but encoding time and computing power consumption increase
Solution Approach 1:
The patent divides the target frame image into multiple local image blocks and embeds watermark information into specific blocks rather than the entire image. This segmentation approach reduces the total number of pixels requiring watermark embedding, thereby decreasing encoding time and computing power consumption while maintaining sufficient watermark coverage through strategic block selection
Solution Approach 2:
The patent applies different watermark embedding strategies to different local image blocks based on their characteristics. By selecting specific blocks for watermark embedding and using adaptive embedding intensities, the system achieves effective watermark coverage without uniformly processing the entire image, thus reducing overall encoding time and computational requirements
2Manufacturing precision
If video resolution is increased, then image quality is improved, but encoding speed decreases and computing power consumption increases
Solution Approach 1:
The patent processes the high-resolution video frame by dividing it into multiple local image blocks, allowing parallel processing of smaller regions. This segmentation enables maintenance of high image quality through precise block-level operations while improving encoding speed by distributing the computational load across multiple smaller processing units
Solution Approach 2:
The patent applies watermark embedding to only certain local image blocks rather than all blocks in the frame. This partial action approach maintains high image quality in critical regions while reducing the total computational effort required, thereby improving encoding speed without sacrificing overall image quality
3Reliability
If watermark embedding is performed on the entire frame, then decoding robustness is improved, but vulnerability to geometric transformations increases
Solution Approach 1:
The patent embeds watermark information across multiple distributed local image blocks rather than concentrating it in one area. This segmentation provides redundancy, as the watermark can be recovered from multiple blocks even if some are affected by geometric transformations, thereby maintaining decoding robustness while reducing vulnerability to localized geometric attacks
Solution Approach 2:
The patent performs geometric correction on local image blocks before watermark embedding by using anchor points to establish proper spatial relationships. This preliminary geometric normalization ensures that watermarks are embedded in a standardized coordinate system, making the watermark more resistant to subsequent geometric transformations and improving overall decoding robustness
Data Source
AI summary
This application provides techniques of processing video frames. The techniques comprise obtaining a target frame of a video; capturing at least one to-be-processed image from the target frame, wherein the at least one to-be-processed image comprises at least one segment of the target frame; generating a plurality of channels by performing channel decomposition on the to-be-processed image, wherein the plurality of channels comprise a first channel and a second channel; embedding watermark information indicative of a watermark into the first channel; embedding location information into the second channel, wherein the location information indicates a location of the watermark in the target frame; generating a target image corresponding to the to-be-processed image based on the first channel comprising the embedded watermark information and the second channel comprising the embedded location information; and generating a watermarked video by embedding the target image into the target frame.