Bottom-of-Basket Item Detection Using Visual Feature Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting Bottom-of-the-Basket (BoB) losses in retail settings, such as using mirrors, video cameras, or special markers, are not effective due to low accuracy and high false alarm rates, leading to cashier indifference and continued merchandise loss.
Innovation Solution
A system and method that extracts scale-invariant visual features from images of shopping carts using SIFT or SURF features, tracks these features across images, and classifies them based on motion and appearance parameters to accurately detect the presence of items on the cart without the need for reflectors or markers, generating alerts when specific criteria are met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If mirrors or video cameras are installed to provide visual inspection of bottom-of-basket, then cashier visibility of items is improved, but cashier workload and complexity of the system increases
Solution Approach 1:
The patent replaces mechanical visual inspection systems (mirrors, video cameras) with an automated image processing system that uses feature extraction, motion detection, and classification algorithms to automatically detect items in the bottom-of-basket, eliminating the need for cashiers to manually monitor additional visual displays
Solution Approach 2:
The system performs self-detection by automatically analyzing images to identify items in the bottom-of-basket without requiring cashier intervention or attention, with the classification algorithm autonomously determining the presence and characteristics of items
2Reliability
If special markers or reflectors are placed on carts to enable detection, then detection capability is improved, but ease of operation and customer experience deteriorates
Solution Approach 1:
The patent extracts and removes the need for special markers or reflectors on carts by using natural visual features of items themselves (edges, corners, textures) for detection, allowing standard carts to be used without modifications while maintaining reliable detection capability
Solution Approach 2:
The system changes the detection approach from looking for specific artificial markers to analyzing natural visual parameters of items (scale-invariant features, motion patterns, appearance characteristics), enabling detection of standard carts with regular merchandise
3Speed
If beam probing is used to detect items in bottom-of-basket, then detection speed is improved, but measurement precision and false alarm rate worsens
Solution Approach 1:
The system uses feedback loops where detected features are classified and validated against multiple criteria (motion parameters, appearance parameters, classification thresholds), with the ability to adjust detection sensitivity and reduce false alarms based on accumulated detection data and pattern recognition
Solution Approach 2:
The system performs preliminary feature extraction and motion detection on image sequences before final classification, pre-processing the visual data to identify potential items and their characteristics, which speeds up the final detection decision while maintaining accuracy through multi-stage verification
4Reliability
If continuous monitoring of cart bottom is required from cashier, then item detection is improved, but productivity and cashier efficiency deteriorates
Solution Approach 1:
The patent substitutes the cashier's manual visual monitoring task with an automated computer vision system that continuously analyzes images of the bottom-of-basket, freeing the cashier to focus on customer service and transaction processing while maintaining reliable item detection
Solution Approach 2:
The system provides continuous automated monitoring of the bottom-of-basket throughout the transaction process, continuously analyzing image sequences to detect items at any stage, ensuring no items are missed while allowing the cashier to maintain normal workflow without intermittent interruptions
Data Source
AI summary
A system and method for detecting the presence of known or unknown objects based on visual features is disclosed. In the preferred embodiment, the system is a checkout system for detecting items of merchandise on a shopping cart. The merchandise checkout system preferably includes a feature extractor for extracting visual features from a plurality of images; a motion detector configured to detect one or more groups of the visual features present in at least two of the plurality of images; a classifier to classify each of said groups of the visual features based on one or more classification criteria, wherein each of the one or more parameters is associated with one of said groups of visual features; and an alarm configured to generate an alert if the one or more parameters for any of said groups of the visual features satisfy one or more classification criteria.