Robust Image Signature for Tamper Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image authentication methods fail to distinguish between allowable modifications, such as lossy compression, and malicious tampering, and are not robust enough to survive high compression ratios or accurately localize tampering in digital images and videos.
Innovation Solution
A progressive and robust signature is generated by splitting the audio-visual signal into blocks, calculating DC-values, assigning blocks to regions with similar DC-values, and generating signature bits based on DC-differences, which allows for adaptable and localized tampering detection with minimal payload.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If classical cryptography is used to authenticate images, then the authenticity validation is straightforward, but any bit change including allowable compression modifications causes complete hash mismatch and false tamper detection
Solution Approach 1:
The image is divided into multiple blocks, and the signature is computed based on aggregated properties (DC-values, edges, moments) of these blocks rather than individual pixels. This segmentation allows the signature to tolerate minor modifications within blocks while still detecting malicious tampering.
Solution Approach 2:
The patent transforms the image from pixel-level data to feature-level data (DC-values, edges, moments, histograms). This parameter transformation makes the signature robust to compression and other allowable modifications that preserve these higher-level features while being sensitive to malicious tampering.
2Adaptability or versatility
If semi-fragile watermarks are embedded for authentication, then tolerance to compression is achieved, but the watermarks cannot survive high compression ratios and fail to provide robust tamper detection
Solution Approach 1:
The signature is computed and embedded in the image before any compression or processing occurs. This preliminary embedding ensures that the signature survives subsequent compression operations, as it is embedded in a robust manner that accounts for future degradation.
Solution Approach 2:
The signature combines multiple image properties (DC-values, edges, moments, histograms, compression invariants) to create a composite authentication mechanism that is more robust than any single property alone, enabling survival through high compression ratios.
3Reliability
If robust watermarks are used to allow correct extraction after tampering, then tamper detection capability is improved, but the payload size is very limited to just tens of bits
Solution Approach 1:
The patent extracts only the essential authentication information into a compact signature, separating it from the full image data. This extraction allows robust tamper detection with minimal payload, as only the critical features (DC-values, edges, moments) are embedded rather than the entire image.
4Measurement precision
If the signature size is increased to provide higher level of protection and accurate tamper localization, then the ability to detect and localize tampering is improved, but the storage and transmission requirements increase significantly
Solution Approach 1:
The image is divided into blocks, and the signature is computed from aggregated block properties. This segmentation enables accurate tamper localization by identifying which specific blocks have been modified, while keeping the overall signature size manageable through efficient representation of block-level features.
Data Source
AI summary
A method, system and computer readable medium for the authenticating of an audio-visual signal, such as digital images or video comprising the generation of a robust image signature with variable size. In a preferred embodiment DC-values of blocks of a digital image are calculated and areas with similar DC-values are merged into regions. The signature is based on said regions and of variable length, depending on the desired localization ability or an allowable signature length. The resulting signature bits are robust to compression and other allowable image operations. The hierarchical solution provides both robustness and tampering localization.