Optical Pattern Recognition for Object Location in Retail
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Retail, logistics, and fulfillment centers face challenges in efficiently locating items on shelves due to similarities in item appearance and lack of adequate information, leading to errors in order fulfillment and customer support, and implementing custom augmented reality (AR) solutions requires significant resources and time.
Innovation Solution
A system using optical pattern recognition on mobile devices captures images of objects with affixed optical patterns, decodes them to retrieve data, and overlays AR graphics to aid users in quickly finding matching items, reducing the need for custom AR solution development and enhancing search accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional barcode scanning methods are used, then object tracking can be achieved, but the process requires specialized scanners and is time-consuming
Solution Approach 1:
The patent replaces specialized laser-based barcode scanners with mobile device cameras and optical pattern recognition algorithms. This substitution enables standard mobile devices to perform scanning functions, significantly improving accessibility and reducing the time required to locate items while maintaining decoding accuracy.
Solution Approach 2:
The system pre-processes and stores optical pattern data in databases before actual scanning operations. By having reference patterns and object information readily available in advance, the system can quickly match captured images against stored data, reducing the time required for object identification and location.
2Ease of operation
If custom augmented reality solutions are implemented, then visual assistance can be provided, but significant resources and time are required for development
Solution Approach 1:
The patent implements a universal optical pattern recognition system that can be integrated into standard mobile devices without requiring custom augmented reality development. The system provides visual assistance through graphic overlays that indicate object locations and provide navigation guidance, leveraging existing mobile device capabilities rather than requiring specialized custom AR solutions.
Solution Approach 2:
The system introduces an intermediary layer of optical pattern recognition that bridges the gap between simple barcode scanning and complex custom AR solutions. By using standardized graphic overlays and visual indicators that work across different mobile devices, the system provides enhanced user guidance without requiring device-specific custom AR development.
3Measurement precision
If items are located using appearance similarity, then visual matching can be performed, but errors increase due to similar-looking items
Solution Approach 1:
The patent employs optical patterns with distinct visual characteristics including color variations, shapes, and configurations that are affixed to items. These patterns serve as unique identifiers that are visually distinct even when items have similar appearances, enabling accurate differentiation and identification. The system uses these visual features to match items against database records, significantly improving identification accuracy and reducing errors in order fulfillment.
Data Source
AI summary
Features described herein generally relate to locating objects in a scene using optical pattern recognition. Images of a scene that include objects are captured and optical patterns associated with the objects are detected and decoded. Code data for the optical patterns can be matched to search parameters matching input search criteria. The images can be displayed along with a graphic that overlays at least one of the images to indicate that an object matching the search criteria is depicted in the scene.