Shopping Cart Camera with Illumination for Item Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cart POS systems using AI object recognition struggle to identify all items perfectly, necessitating the use of barcode scanners, which can be confusing for customers and difficult to position correctly, as the terminal with a barcode scanner is often outside the basket.
Innovation Solution
An item registration apparatus equipped with a camera and a lighting unit that uses object recognition initially and switches to identification code recognition with illumination guidance when an item is not recognized, eliminating the need for a separate barcode scanner by indicating the reading position through illumination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a separate barcode scanner is attached to the shopping cart for item identification, then item recognition accuracy is improved, but device complexity and customer confusion increase
Solution Approach 1:
The patent combines the barcode scanning function with the existing camera-based object recognition system into a single integrated terminal. The camera serves dual purposes: capturing images for AI-based object recognition and capturing images for barcode recognition when needed. This eliminates the need for a separate barcode scanner attachment, reducing device complexity while maintaining item recognition accuracy.
Solution Approach 2:
The camera in the shopping cart terminal is designed to perform multiple functions: it captures images for AI-based object recognition of items placed in the cart, and it also captures images for barcode recognition when the object recognition system cannot identify an item. This multi-functional design eliminates the need for separate dedicated devices for each recognition method.
2Measurement precision
If a barcode scanner is installed outside the basket for item reading, then item identification capability is improved, but ease of operation deteriorates due to difficult positioning
Solution Approach 1:
The barcode scanning function is merged with the camera positioned inside or near the basket, rather than being installed separately outside the basket. This integration allows customers to simply place items in the basket and the camera to automatically capture both the item for object recognition and the barcode if needed, eliminating the need for customers to manually position items in front of a separate scanner.
Solution Approach 2:
The system performs automatic item recognition through AI object recognition without requiring customer action. When object recognition fails, the system automatically switches to barcode recognition mode, and the camera automatically captures the barcode. The illumination unit automatically activates to guide the customer, creating a self-service system that reduces operational complexity for customers.
3Ease of operation
If AI object recognition is used alone for item registration, then ease of operation is improved, but measurement precision deteriorates due to incomplete item identification
Solution Approach 1:
The system uses a feedback mechanism where the AI object recognition process continuously monitors item placement in the cart. When an item is placed, the system attempts object recognition first. If recognition fails or is uncertain, the system automatically switches to barcode recognition mode, capturing the barcode through the camera and using the illumination unit to guide proper positioning. This feedback loop ensures high identification accuracy while maintaining ease of operation.
Solution Approach 2:
The system performs preliminary AI object recognition on items placed in the cart before attempting barcode scanning. This preliminary action allows the system to identify most items automatically without requiring barcode scanning. Only when the preliminary object recognition fails does the system proceed to barcode recognition, optimizing the process by using the simpler method first when possible.
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 simplifies the item registration process by providing intuitive guidance for customers to read identification codes, reducing confusion and effort, while maintaining efficient item recognition without the need for a dedicated barcode reader, thus enhancing user experience and operational efficiency.
Implementation Method 1
control the camera to capture a first image using first settings, perform an object recognition process on the first image
Implementation Method 2
control the lamp to turn on and control the camera to capture a second image using second settings
Data Source
AI summary
An item registration apparatus for registering an item for purchase in a store, includes a camera, a lamp disposed adjacent to the camera, and a processor configured to: control the camera to capture a first image using first settings, perform an object recognition process on the image, and determine whether a package of an item is in the image, upon determining that a package of an item is in the image, register the item for purchase, upon determining that a package of an item is not in the image, control the lamp to turn on and control the camera to capture a second image using second settings, perform a code recognition process on the second image, and determine whether a code of an item is in the second image, and upon determining that a code of an item is in the second image, register the item using the code.