Check Segmentation for Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods fail to effectively detect and deter check re-presentment fraud, which poses a liability for financial services institutions, as they rely on matching deposit amounts and MICR lines, making it difficult to identify re-presented checks.
Innovation Solution
The method involves separating a check into portions, either by cutting or tearing, and generating digital images of these portions, which are then processed by the institution to create a combined image or using holes/marks to align and process the check, making it harder for users to re-present the check and easier for institutions to detect fraud.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional matching techniques (deposit amount and MICR line) are used to detect re-presentment, then the detection process is simple, but the detection effectiveness is poor and fraud liability remains high
Solution Approach 1:
The check is divided into multiple portions (e.g., left and right sections) before imaging. Each portion contains unique visual features that can be independently identified. This segmentation prevents re-presentment because the system can detect that the portions have already been processed and combined, making fraud detection more reliable without requiring complex additional hardware.
2Reliability
If checks are separated into portions and digital images are generated, then re-presentment detection improves, but the processing complexity increases
Solution Approach 1:
Visual features such as markers, patterns, or distinctive elements are embedded in the check portions during the initial separation stage. These pre-placed features enable the imaging system to automatically identify and track portions without requiring complex real-time analysis, thus improving detection accuracy while keeping processing complexity manageable.
3Ease of operation
If multiple digital images of check portions are combined, then a complete check image is formed for processing, but the risk of re-presentment increases if not properly tracked
Solution Approach 1:
The system incorporates feedback mechanisms that track the status of check portions through the imaging and combination process. Each portion is identified using its unique visual features, and the system maintains a record of processed portions. This feedback loop ensures that re-presented checks are detected and rejected, reducing fraud risk while maintaining ease of operation for legitimate deposits.
Data Source
AI summary
A user or a device may separate a check into two or more portions prior to generating a digital image of the check for remote deposit of the check. The user or a device may separate the check by cutting or tearing the check. After separating the check into the portions, the user may generate a digital image of the portions of the check using a scanner for example. The digital image may be transmitted to an institution for deposit of the check. The institution may retrieve the images of the portions of the check and generate an image of the check based on the portions, by combining the images of the portions for example. The image of the check that may be generated based on the images of the portions may be processed for deposit.