3D Triangulation Item Dimensioning for POS Fraud Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point of sale systems relying solely on optical code scanners are vulnerable to fraudulent transactions due to misidentification of items, as they do not account for physical dimensions, leading to increased computation time and fraud.
Innovation Solution
A method and apparatus using multiple cameras to capture images from different angles, identifying corners and projections onto reference planes to determine the dimensions of an unknown item, thereby reducing the universe of known items to compare against and enhancing identification speed and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If optical code scanning is used for item identification, then processing speed is improved, but fraud vulnerability increases due to reliance solely on codes
Solution Approach 1:
The patent combines optical code scanning with image-based dimensional measurement systems. The point of sale terminal integrates both code reading capabilities and 3D imaging sensors to simultaneously perform code identification and physical characteristic verification, creating a multi-modal identification system that reduces fraud while maintaining speed
Solution Approach 2:
The system uses image processing to extract dimensional information and compares it against expected dimensions for the identified item. This feedback mechanism verifies the authenticity of the item by checking whether physical measurements match the product database, thereby detecting fraudulent substitutions
2Reliability
If image processing is used to identify items by physical characteristics, then fraud mitigation is improved, but computation time increases
Solution Approach 1:
The system pre-processes images to quickly extract key dimensional features such as height, width, and depth by identifying corners and edges. This preliminary feature extraction reduces the complexity of subsequent comparison operations against the product database, enabling fast fraud detection without full image analysis
Solution Approach 2:
The patent extracts only the essential dimensional characteristics (height, width, depth) from the full image data by detecting corners and projecting onto reference planes. This selective extraction of critical features reduces computational load while maintaining sufficient accuracy for fraud detection
3Reliability
If full item comparison is performed against the entire known items database, then identification accuracy is improved, but processing time increases
Solution Approach 1:
The patent segments the item identification process into hierarchical stages: first comparing dimensional characteristics (height, width, depth), then shape features, and finally detailed image matching. This segmentation allows the system to quickly eliminate large portions of the database using coarse dimensional filters before performing more detailed comparisons on a reduced subset
Data Source
AI summary
Methods and apparatus are provided for determining the dimensions of an unknown item using multiple images captured by different cameras. The dimensions include the height, length and width of the item. The dimensions are used to reduce the universe of possible candidates that would identify the unknown item. The methods use multiple reference planes and projects elements of the item onto the one or more of the reference planes to determine the coordinates of the elements of the item. Once the coordinates are determined, the dimensions of the items are determined.