Video Watermark Detection Accuracy Under Compression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for detecting embedded watermarks in video streams are limited in accuracy due to compression errors, especially when identifying which palette was used in low variance or low edge areas, making it difficult to trace unauthorized distribution of content.

Innovation Solution

A method and apparatus that embeds watermarks by replacing pixels with low variance or low edge areas using a fixed color value from one of two palettes closest to their original average, and detects the watermark by determining which palette was used, incorporating error correction to overcome compression-induced errors.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If watermark detection is performed on compressed video streams, then the ability to trace unauthorized distribution is improved, but compression errors reduce detection accuracy

Engineering Contradiction:
Improvewatermark detection accuracyVSAvoidcompression errors
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent applies error correction coding during the watermark embedding process, preparing corrected data in advance before compression occurs. This preliminary error correction ensures that even when compression introduces errors, the watermark can be accurately detected after compression.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The detection process uses confidence thresholding where the system evaluates the strength of watermark detection signals and compares them against predetermined thresholds. This feedback mechanism allows the system to adjust its detection sensitivity and maintain accuracy despite compression-induced variations.

Inventive Principle:
Principle #23Feedback

2Object-generated harmful factors

If pixel replacement is used to embed watermarks in low variance areas, then the watermark becomes imperceptible, but detection accuracy decreases due to compression errors

Engineering Contradiction:
Improveimperceptibility of watermarkVSAvoidpalette identification accuracy
Core Design Contradiction:
Object-generated harmful factorsVSMeasurement precision

Solution Approach 1:

The patent changes the parameter space by using error correction coded data instead of direct pixel values. By embedding corrected data in the low variance areas and using threshold-based detection, the system maintains both imperceptibility and detection accuracy despite compression errors.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

Error correction coding is applied beforehand to cushion against the harmful effects of compression errors. This pre-prepared correction mechanism ensures that even when compression distorts the embedded watermark data, the original information can be recovered accurately during detection.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentEP3477579B1Method and apparatus for detecting an embedded watermark
Publication Date: 2023.07.05 SYNAMEDIA LTD
  • EP3477579B1 patent drawingFigure 1
  • EP3477579B1 patent drawingFigure 2
  • EP3477579B1 patent drawingFigure 3

AI summary

In one embodiment, a watermark is embedded in a sequence of video frames, for each one of the video frames, a set of N rectangular patches, two palettes of pixel values in a selected color space, such that the two palettes are denoted as P0 and P1, and P0 ≠ P1, and in each one of the N patches a processor which chooses one of P0 and P1 according to a value of a payload bit to be encoded, a calculator which calculates one of a variance of pixel values in the frame, and edginess for the pixel values in the frame, thereby determining a calculated value, and a processor which determines if the calculated value is beneath a given threshold value, if the calculated value is beneath the given threshold value the processor is operative to calculate a function of the pixel values, the closest value to a result of the calculation of the function in the chosen palette is chosen, and a replacer which replaces the pixel values within the patch with the chosen closest value. Related systems, apparatus and methods are also described.