Image-Based Basket Item Volume Estimation for Hidden-Item Counting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for personal item scanning in retail environments struggle to accurately count items in shopping baskets due to hidden items, leading to difficulties in detecting fraudulent behaviors and errors.
Innovation Solution
An information processing system that utilizes object detection algorithms to analyze images from monitoring cameras, correct for rotational biases, and estimate item volumes based on positional outlines to accurately determine the number of items in a basket.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a monitoring camera system is used to detect fraudulent behaviors, then the ability to sense suspicious behaviors is improved, but the accuracy of item counting remains insufficient due to hidden items
Solution Approach 1:
The system segments the item counting process into multiple stages: first detecting the shopping basket region, then detecting individual item regions within it, and finally calculating volumes based on spatial relationships. This segmentation allows the system to handle hidden items by processing visible portions and inferring the rest through systematic division of the detection task.
Solution Approach 2:
The system transitions from two-dimensional image detection to three-dimensional volume estimation by calculating item volumes based on positional outlines and spatial relationships. This dimensional transformation enables more accurate counting even when items are partially hidden, as the system infers hidden portions through volumetric analysis of visible boundaries.
2Ease of operation
If simple item detection is used in self-scan systems, then the ease of operation is improved, but the accuracy of detecting fraudulent behaviors deteriorates
Solution Approach 1:
The system introduces an intermediary processing layer that bridges simple customer scanning and accurate fraud detection. By automatically analyzing shopping basket images, detecting item regions, and calculating volumes to verify item counts, the system maintains ease of self-scan operation while achieving accurate fraud detection through this intermediate verification step.
3Reliability
If video analysis is used to specify item numbers, then the detection capability is improved, but the complexity of the system increases
Solution Approach 1:
The system replaces complex mechanical or manual verification processes with automated image processing and computer vision algorithms. By using software-based analysis to detect item regions, calculate volumes, and verify counts, the system achieves reliable fraud detection while keeping the physical system relatively simple, substituting computational complexity for mechanical complexity.
Data Source
AI summary
A non-transitory computer-readable recording medium stores therein an information processing program that causes a computer to execute a process including, acquiring an image of an object storing items that is captured, extracting an object region of the object contained in the image by analyzing the image acquired, extracting one or more item regions contained in the extracted object region by analyzing the image acquired, estimating an item volume of the items stored in the object based on a position of an outline of the item region with respect to a position of the extracted object region, and specifying a number of items stored in the object based on the specified item volume.