Predictive Watermark Detection Model for Image Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for decoding visually imperceptible watermarks from images are computationally expensive and inefficient, as they often attempt to decode images that do not contain watermarks or have corrupted encodings, leading to unsuccessful decoding processes.

Innovation Solution

A predictive model is employed to analyze features of an image to predict whether it contains a watermark, determining if decoding attempts should be made, thereby saving computational resources by identifying images without watermarks or with undecodable encodings.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If decoding processes are attempted on all images, then watermark detection coverage is improved, but computational time and expense increase significantly

Engineering Contradiction:
Improvewatermark detection coverageVSAvoidcomputational time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent applies preliminary action by implementing a predictive model that analyzes image features before the decoding process. This prediction step determines whether an image contains a watermark, allowing the system to prepare and prioritize decoding operations. By performing this preliminary analysis, the system avoids wasting computational resources on images without watermarks while ensuring that images with watermarks are processed efficiently.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the watermark detection process into two distinct stages: a prediction stage that analyzes image features to determine watermark presence, and a decoding stage that processes only those images predicted to contain watermarks. This segmentation separates the computational workload, allowing the prediction model to filter out unnecessary decoding operations and significantly reduce overall computational time and expense.

Inventive Principle:
Principle #1Segmentation

2Reliability

If decoding processes are attempted on all images, then watermark detection completeness is improved, but computational expense increases significantly

Engineering Contradiction:
Improvewatermark detection completenessVSAvoidcomputational expense
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent implements preliminary action by using a prediction model to assess images before decoding. This preliminary step identifies which images are likely to contain watermarks based on feature analysis, allowing the system to allocate computational resources efficiently. The prediction model serves as a gatekeeper that prevents expensive decoding operations on images without watermarks, thereby reducing overall computational expense while maintaining detection completeness.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the detection workflow into a low-cost prediction phase and a high-cost decoding phase. The prediction model processes all images to identify candidates, while the decoding process is applied only to the subset of images predicted to contain watermarks. This segmentation dramatically reduces computational expense by eliminating unnecessary decoding operations while preserving detection completeness through the two-stage approach.

Inventive Principle:
Principle #1Segmentation

3Productivity

If predictive model analysis is performed, then decoding efficiency is improved, but device complexity increases

Engineering Contradiction:
Improvedecoding efficiencyVSAvoidsystem complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary predictive model that sits between image input and the decoding process. This intermediary component analyzes image features and provides predictions that guide subsequent decoding operations. By inserting this intermediary layer, the system achieves higher decoding efficiency through intelligent filtering, while the added complexity is managed by using a relatively simple machine learning model that can be trained offline and deployed with minimal overhead.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3815034B1Detecting semi-transparent image watermarks
Publication Date: 2024.12.11 GOOGLE LLC
  • EP3815034B1 patent drawingFigure 1
  • EP3815034B1 patent drawingFigure 2A~2C
  • EP3815034B1 patent drawingFigure 2D~2G

AI summary

Systems, methods, devices, and other techniques for detecting, with a predictive model, the presence of a watermarking image in a captured source image. The techniques can involve operations that include obtaining a first image, where the first image is a scaled version of the captured source image. Colors of pixels in the first image are analyzed to determine encoded values of possibly encoded pixels in the first image, and patterns of encoded values of the possibly encoded pixels in the first image are analyzed to determine possibly encoded regions of the first image. Values for features of the possibly encoded regions of the first image are determined, and a predictive model processes the feature values to generate a prediction indicative of whether a watermarking image is encoded in the first image. The prediction is stored and provided to a controller to make a decision regarding invocation of a decoding process.