Item Identification Using Physical Attribute Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point of sale systems face challenges in identifying items efficiently, particularly due to fraudulent methods involving optical code manipulation and issues with damaged or obscured codes, leading to increased transaction times and data storage requirements when relying on image analysis.
Innovation Solution
Implementing a method that uses a scanner to measure and process a set of physical attributes related to an item, reducing a large universe of possible identifications to a smaller set using aggregate attribute data records, allowing for rapid identification without the need for extensive data transmission or processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If image analysis is used to identify items by appearance, then fraud is reduced and identification capability is improved, but data storage requirements and processing time increase significantly
Solution Approach 1:
The patent segments the identification process into two distinct phases: first capturing physical attributes (weight, dimensions, color, shape) that can be quickly measured, then using these segmented attributes to query a database and reduce the universe of possible identifications from 100,000+ items to a small subset, finally performing detailed image analysis only on this reduced set. This segmentation eliminates the need to analyze full images for all items while maintaining high identification accuracy.
Solution Approach 2:
The patent extracts only the essential physical attributes (weight, dimensions, color, shape) from the complete item description, storing and using only these extracted features for initial identification. This extraction approach reduces data storage from gigabytes of full image data to minimal attribute data, while still enabling rapid fraud prevention and item identification.
2Measurement precision
If a large database with detailed item information is used, then identification accuracy is improved, but data transmission and processing time increase
Solution Approach 1:
The patent extracts and stores only the essential physical attributes (weight, dimensions, color, shape) from complete item descriptions in the database. This extraction reduces database record size from 10,000-100,000 bytes to minimal attribute data, enabling rapid querying and identification without transmitting large amounts of data between the point of sale terminal and central server.
Solution Approach 2:
The identification process is segmented into attribute-based filtering followed by detailed verification. The database is organized to support this segmentation, with physical attributes stored separately and used for rapid initial filtering, allowing the system to achieve high precision identification with minimal data transmission.
3Productivity
If optical code scanning is used for item identification, then identification speed is improved, but vulnerability to fraud and unreadable codes increases
Solution Approach 1:
The patent implements a multi-functional identification system that can operate in multiple modes: optical code scanning for rapid identification when codes are readable, physical attribute measurement for fraud prevention, and image analysis for items with damaged or missing codes. This universal approach maintains high transaction speed while eliminating vulnerability to code manipulation fraud and handling unreadable codes.
Solution Approach 2:
The patent introduces physical attribute measurements (weight, dimensions, color, shape) as an intermediary between optical code scanning and final item identification. This intermediary layer provides a verification mechanism that detects fraud (such as weight mismatches) and enables identification when optical codes are unreadable or manipulated, while maintaining rapid transaction processing.
Data Source
AI summary
Methods and apparatus are provided for fast item identification. A plurality of sensors capture data for an unknown item that is moved past the sensors. The data is processed to produce a plurality of physical attributes related to the unknown item. The physical attributes are used to search a database of physical attributes for a large number of known items where the unknown item is one of the known items. A small set of known items are selected where the physical attributes of the known items match the physical attributes of the unknown item. Further processing of the selected set of known items identifies the unknown item as one of the set of known items.