Value Document Forgery Detection via Maximum Flow Network Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting adhesive strips and composed forgeries in value documents face challenges due to thickness measurement fluctuations and limitations in distinguishing thin strips from document thickness profiles, and struggle to accurately identify manipulated or forged documents.
Innovation Solution
A method utilizing spatially resolved measurements to form a two-dimensional network of nodes, calculating capacitance values between nodes, and determining the maximum possible flow through the network to classify documents as authentic or forged, effectively identifying continuous objects like adhesive strips or separation lines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If thickness measurement methods are used to detect adhesive strips, then detection capability is provided, but measurement precision deteriorates due to thickness fluctuations and interference from document thickness profiles
Solution Approach 1:
The document surface is divided into a two-dimensional network of nodes, where each node represents a measurement point. This segmentation allows the system to analyze local variations in measured values and distinguish adhesive strips from normal thickness profiles by comparing patterns across multiple discrete points rather than relying on overall thickness measurements.
Solution Approach 2:
The invention transitions from one-dimensional thickness measurements to two-dimensional spatial distribution analysis of measured values. By creating a node network that maps measurements across the document surface and analyzing flow patterns through this network, the system can identify continuous objects like adhesive strips based on their spatial extent and connectivity, overcoming the limitations of simple thickness thresholds.
2Reliability
If current tape detection methods are used, then some tape can be detected, but detection reliability deteriorates when tape is very thin or when document thickness profile is pronounced
Solution Approach 1:
The system pre-establishes a two-dimensional node network across the document surface before detection, assigning expected measured values to each node based on document characteristics. This preliminary framework enables the system to compare actual measurements against expected patterns, making detection more reliable for thin tapes and documents with pronounced thickness profiles by identifying deviations from the established baseline.
Solution Approach 2:
The system uses the calculated maximum flow through the node network as feedback to assess the presence of continuous objects. By comparing the actual flow pattern against expected flow patterns for authentic documents, the system can reliably detect thin adhesive strips and composed forgeries, adjusting detection sensitivity based on the overall document measurement pattern rather than relying on fixed thresholds.
3Difficulty of detecting and measuring
If spatially resolved optical measurement methods are used to detect composed forgeries, then detection capability is provided, but device complexity increases
Solution Approach 1:
The patent introduces a mathematical intermediary - the maximum flow calculation through a node network - that translates complex spatial measurement data into a single detectable parameter. Instead of directly analyzing complex optical measurement patterns, the system uses flow network theory as an intermediary to identify continuous objects, simplifying the detection process while maintaining high capability for detecting composed forgeries.
Solution Approach 2:
The invention transforms the detection parameter from raw measured values at individual points to the maximum flow through the entire node network. This parameter transformation converts complex spatial distribution data into a single scalar value that directly indicates the presence of continuous objects, reducing device complexity while preserving detection capability for composed forgeries.
Data Source
Figure 1~6
Figure 3a~3d
Figure 4
AI summary
A method for checking a value document is described in which measured values from the value document are detected in spatially resolved fashion. The measured value detected at the respective measurement point is associated with a node that corresponds to said measurement point, and a two-dimensional network of nodes is formed therefrom. A network is formed from the two-dimensional network of the nodes and also a source node and a sink node. On the basis of the maximum possible flow through the network, the value document is classified as suspected forgery or not suspected forgery. The maximum possible flow through the network is a measure of the probability of the value document having, in a direction transverse to the direction of the network, a continuous object that indicates manipulation of the value document, such as an adhesive strip or a separating line from a composed forgery.