Multi-Zone Image Search for Check Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated processing schemes for note payables, such as checks, lack flexibility in searching multiple zones for information, leading to inefficiencies and increased risk of fraud.
Innovation Solution
A method and system that search multiple zones of an image file for anchors or keywords, compare the obtained information with corresponding information from an issue file, and provide an indication for rejection if the information does not match, using optical character recognition and loose matching techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If automated processing schemes search for information in only one area of a note payable, then the processing speed is maintained, but the flexibility and thoroughness of fraud detection is reduced
Solution Approach 1:
The image file is divided into multiple search zones (e.g., first zone, second zone, third zone) where each zone is searched independently for anchors or keywords. This segmentation allows the system to search multiple areas without processing the entire image sequentially, thereby improving flexibility while maintaining processing efficiency.
Solution Approach 2:
The system transitions from searching a single area to searching multiple zones across different spatial dimensions of the image file. By utilizing the spatial distribution of zones, the system achieves comprehensive fraud detection without sacrificing processing speed, as each zone can be searched in parallel or with optimized scanning patterns.
2Reliability
If multiple zones are searched for information, then the flexibility and fraud detection capability is improved, but the processing time and complexity increase
Solution Approach 1:
The system performs preliminary actions by pre-defining multiple search zones and pre-identifying potential anchors or keywords before actual processing. This preparation allows the system to quickly search each zone without performing time-consuming analysis during the main processing phase, thereby reducing overall processing time while maintaining high fraud detection accuracy.
Solution Approach 2:
The system maintains continuous useful action by efficiently transitioning between search zones without idle time. Once one zone is processed, the system immediately proceeds to the next zone, ensuring that the entire image file is searched thoroughly and continuously. This continuous processing approach minimizes delays and maintains high throughput while detecting fraud across all zones.
3Measurement precision
If strict matching is used to compare information, then the accuracy of fraud detection is improved, but the number of false rejections increases
Solution Approach 1:
The system changes the matching parameter from strict equality to loose matching that allows for predetermined variations (e.g., ignoring case differences, allowing for typos, or accepting synonyms). This parameter change enables the system to maintain high accuracy in detecting actual fraud while reducing false rejections of legitimate checks, thereby improving overall processing throughput.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the flexibility and efficiency of processing note payables, reducing the risk of fraud by allowing multiple searches within an image file and comparing information with issue files, thereby improving the reliability of check processing.
Implementation Method 1
searching the image file in multiple zones for an anchor or a keyword
Data Source
AI summary
Methods, apparatuses, systems, and tangible computer readable media for processing note payables for fraud by searching an image file of a note payable for information in multiple zones on the image file and comparing the obtained information to corresponding information in the issue file that is associated with the note payable. The note payable may be a check and searching of the image file may be done by optical character recognition. However, a user may wish to search an image file of a note payable in multiple zones for desired information.