Self-Checkout Motion Detection for Automatic Product Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing self-checkout systems require significant customer intervention, leading to delays and inconvenience in high-throughput sales environments due to the need to register each product individually, especially those without scannable identifiers, such as weight-dependent items.
Innovation Solution
A self-checkout device equipped with cameras to capture video footage, a motion detection module, and an appearance interpretation module to automatically identify products, along with a billing module to calculate the total bill, reducing the need for manual product registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If existing self-checkout systems are used, then automation is achieved, but customer intervention is still required leading to delays
Solution Approach 1:
The system enables products to register themselves by being placed on the detection plate. The camera captures images, the processor identifies products through image recognition, and registration occurs automatically without customer interaction. This self-service mechanism eliminates the time customers spend manually searching for and selecting products.
Solution Approach 2:
The patent replaces the mechanical/manual process of product registration with an optical system. Cameras capture visual information, and image processing algorithms automatically identify and register products. This substitution of mechanical customer action with optical-digital recognition system dramatically reduces checkout time.
2Measurement precision
If manual product registration is required, then product identification accuracy is maintained, but processing speed decreases
Solution Approach 1:
Manual visual inspection and manual registration are replaced with automated image recognition technology. The camera captures product images, and the processor uses image processing algorithms to automatically identify products with high accuracy while simultaneously achieving fast processing speeds, resolving the contradiction between accuracy and speed.
Solution Approach 2:
The system creates visual copies (images) of products and processes these copies digitally to identify and register products. This copying approach allows parallel processing of multiple products simultaneously, maintaining identification accuracy while dramatically increasing registration speed compared to sequential manual processing.
3Loss of information
If individual product registration is required, then product details are accurately captured, but multiple products cause delays
Solution Approach 1:
The system merges the registration process for multiple products into a single simultaneous operation. The camera captures all products on the detection plate in one image, the processor identifies multiple products concurrently through image recognition, and all products are registered together. This merging eliminates the sequential delays of individual registration while maintaining complete product detail accuracy.
Solution Approach 2:
The system maintains continuous operation by processing multiple products simultaneously without interruption. Once products are placed on the detection plate, the camera continuously captures images and the processor continuously identifies and registers products in real-time, eliminating the start-stop nature of individual registration and reducing total checkout time.
4Reliability
If scannable identifiers are required, then product identification is reliable, but products without identifiers cannot be registered
Solution Approach 1:
The system replaces barcode/RFID scanning with optical image recognition. The camera captures visual information and the processor uses image processing algorithms to identify products based on their visual appearance rather than electronic identifiers. This substitution enables reliable identification of products without scannable identifiers while maintaining system reliability through advanced image recognition technology.
Solution Approach 2:
The image recognition system provides universal product identification capability that works with all products regardless of whether they have scannable identifiers. The system can identify products with barcodes, RFID tags, or no identifiers at all, making the self-checkout system universally applicable to diverse product types and significantly increasing adaptability.
Data Source
AI summary
A self-checkout device and a method implemented by the self-checkout device is provided. The self-checkout device includes a detection plate to allow placement of product(s) thereon, and one or more cameras positioned to have a Field-of-View encompassing the detection plate to provide a video footage. A motion detection module detects presence of motion in the video footage. A sequence selection module selects a sequence of video frames over a time interval corresponding to the detection of the presence of motion in the video footage. An appearance interpretation module registers one or more products present in the sequence of video frames and a billing module fetches prices of the registered one or more products. Total bill is generated based on the fetched prices, and a payment is processed for the total bill.