Shopping Cart Product Identification via Optical Recognition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing self-checkout systems in retail environments face challenges such as high costs for RFID labels, difficulties in identifying products with colorful appearances, and inefficiencies in image recognition processes, particularly when dealing with large product databases, which can lead to inaccuracies and complications for customers and merchants.

Innovation Solution

A system and method utilizing optical cameras mounted on shopping carts connected to a CPU with image processing software that compares product images with a database of signature data from multiple angles, incorporating barcode and text recognition, and optional infrared cameras for temperature analysis, to enable real-time automatic product identification and checkout.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If comprehensive image comparison is performed against large product databases (30,000 to 100,000 items), then recognition accuracy improves, but processing time increases beyond acceptable limits

Engineering Contradiction:
Improverecognition accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the large product database into multiple smaller category-specific databases (e.g., fresh produce, packaged goods, beverages). When a product image is captured, the system first identifies the product category and then performs image comparison only within that specific category database, reducing the search space from 100,000 items to a manageable subset while maintaining high recognition accuracy

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary categorization of products based on visual characteristics (color, shape, packaging type) before conducting detailed image comparison. This preliminary action narrows down the potential matches and enables faster processing by pre-filtering the database based on category attributes

Inventive Principle:
Principle #10Preliminary action

2Reliability

If image recognition is performed during shopping to ensure high accuracy (99% or more), then product identification reliability improves, but the shopping experience is disrupted due to time delays

Engineering Contradiction:
Improveproduct identification reliabilityVSAvoidshopping experience
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The shopping cart system operates autonomously to identify and track products without requiring customer intervention. The camera continuously captures images and the system automatically processes them, allowing customers to shop naturally without being interrupted by scanning or identification procedures

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The image capture and processing occurs continuously as the customer shops, rather than requiring pauses or specific actions. The system maintains a running tally of products in the cart by continuously analyzing images from the camera, ensuring reliable identification without disrupting the shopping flow

Inventive Principle:
Principle #20Continuity of useful action

3Measurement precision

If barcode scanning is required at self-checkout terminals, then product identification accuracy improves, but accessibility deteriorates for elderly or physically handicapped customers

Engineering Contradiction:
Improveproduct identification accuracyVSAvoidcustomer accessibility
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent replaces the mechanical barcode scanning process with an optical image recognition system using computer vision algorithms. The camera captures images of products and the system automatically identifies them through pattern recognition, eliminating the need for customers to physically manipulate products or locate barcodes

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system performs product identification automatically without requiring customer action. The camera and processing system work autonomously to identify products as they are placed in the cart, removing the burden of manual barcode scanning from customers while maintaining high identification accuracy

Inventive Principle:
Principle #25Self-service

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

This solution provides reliable, efficient, and accurate product identification in real-time, minimizing human intervention and improving the shopping experience by ensuring high recognition accuracy and reducing operational costs, while allowing for wireless payment options.

Implementation Method 1

an optical camera (15) mounted to a shopping cart (1)

Methodology Applied
Scientific EffectLight reflection: Reflection

Implementation Method 2

optional infrared cameras for temperature analysis

Methodology Applied
Scientific EffectThermal radiation: Thermal Radiation

Data Source

PatentEP3262562B1System and method for the identification of products in a shopping cart
Publication Date: 2022.10.19 PENTLAND FIRTH SOFTWARE
  • EP3262562B1 patent drawingFigure 1
  • EP3262562B1 patent drawingFigure 2
  • EP3262562B1 patent drawingFigure 3~4

AI summary

A system for the identification of products in a shopping cart in a retail environment comprises: a computer comprising a database with product specific data related to products from an assortment of products in a retail environment; an optical camera connected to the computer, for monitoring the shopping cart; a software stored in a memory of the computer, for identifying products in the shopping cart, which comprises an image recognition software adapted for discerning between different products in the shopping cart, and to determine a product and image areas attributed to it, and to calculate product specific information from the optical data of the image areas; an optional barcode scanner, which may be integrated in the software; wherein the product specific data in the data base comprises signature data derived from previously obtained optical images of the assortment of products, and wherein the software is adapted to identify products in the shopping cart by comparing obtained product specific information of a product in the shopping cart, and wherein the software is adapted to use at least one auxiliary information from at least one further sensing device, in order to improve the accuracy and/or speed of identification of products in the data base. Further, a respective method and computer program product is provided.