Multi-Modal Sensor Fusion for Transaction Fraud Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current transaction monitoring systems face challenges in detecting fraud and abnormal activity due to the unpredictable nature of human behavior and the limitations of human monitoring, which is expensive and prone to errors, and existing solutions like camera systems and data mining are not scalable or effective.
Innovation Solution
A multi-modal sensor system that combines image and depth data to detect abnormal activity by fusing sensor data from visual and depth information, allowing for the identification of patterns not discernible from image data alone, and generates alerts for abnormal behavior.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human monitoring is used to detect fraud and abnormal activity, then detection capability is provided, but cost increases and coverage area is limited
Solution Approach 1:
The patent uses multiple cameras to capture images from different viewpoints, creating visual copies of the transaction area. These multiple views are then presented to a remote guard on television screens, effectively extending the human monitor's coverage area without requiring multiple physical guards. The system creates a visual replica of the monitored space that can be viewed remotely.
Solution Approach 2:
The patent introduces television screens and video transmission as intermediaries between the transaction area and the human guard. This allows the guard to monitor multiple areas simultaneously through the mediated visual feed, expanding coverage without the guard physically being present in multiple locations.
2Reliability
If human monitoring is used to detect fraud, then detection capability is provided, but the system becomes vulnerable to bribery and human error
Solution Approach 1:
The patent replaces direct human monitoring with an automated computer-based monitoring system that uses image processing and data mining algorithms. This mechanical/automated system is not susceptible to bribery or human error, while still providing fraud detection capability. The system objectively analyzes transaction data and visual feeds without human intervention in the detection process.
3Area of stationary object
If traditional camera systems are used to expand monitoring coverage, then area coverage is improved, but scalability and effectiveness remain limited
Solution Approach 1:
The patent creates a multi-functional system that simultaneously performs multiple tasks: capturing visual data from multiple viewpoints, recording transaction data, performing real-time image processing, conducting data mining analysis, and presenting information to guards. This universal system handles both visual monitoring and analytical detection functions in one integrated platform, improving scalability and effectiveness compared to separate traditional systems.
Solution Approach 2:
The patent implements continuous monitoring and analysis where cameras continuously capture images, the system continuously processes visual data and transaction data in real-time, and guards continuously receive updated information. This continuous operation without interruption enhances the system's effectiveness and ability to detect fraud as it occurs, rather than through periodic or batch processing.
Data Source
AI summary
A transaction monitoring system is provided for monitoring transactions in a variety of transaction locations such a retail locations or gambling or other locations. The transaction monitoring system uses a multi-modal sensor an comprises a system controller that uses multi-modal sensor data to detect abnormal activity. The transaction monitoring system may capture and/or use image data and depth data. The transaction monitoring system may be used to detect theft or fraud for example by determining when there is an unscanned item, an occluded item, an incorrectly scanned item, unmatched plurality or incorrect capital exchange. To this end, the transaction monitoring system may also receive transaction input such as barcode data, or capital exchange data. Also provided is a transaction monitoring method and a system controller for a transaction monitoring system.