Shelf Image Depth Analysis for Outlier Product Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual inspection of product storage facilities is time-consuming and increases operational costs due to the large number of shelves and products, necessitating a more efficient inventory management system.
Innovation Solution
A system utilizing a trained machine learning model to process captured images of product storage facilities, determining object coordinates, depth values, and clustering objects to automate inventory updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual inspection is used to inventory products, then workers can identify stocked and out-of-stock products, but the process becomes very time-consuming and increases operational costs
Solution Approach 1:
The patent replaces the manual mechanical inspection process with an automated image processing system. Image capture devices photograph product storage areas, and machine learning models automatically analyze these images to identify products, their locations, and stock status, eliminating the need for manual visual inspection while maintaining accuracy.
Solution Approach 2:
The system creates digital copies (images) of the physical product storage areas. These image copies are then processed by machine learning models to extract inventory information, allowing the system to analyze the stored state without physically interacting with or manually examining each product.
2Quantity of substance
If manual inspection is deployed across hundreds of shelves and thousands of products, then complete inventory coverage is achieved, but operational costs significantly increase
Solution Approach 1:
The machine learning model serves multiple functions: it identifies products, determines their locations on shelves, detects stock status, and can recognize outliers. This multi-functional approach allows a single automated system to replace multiple manual inspection tasks, reducing operational costs while maintaining comprehensive inventory coverage.
Solution Approach 2:
The system enables self-service inventory management where the image processing and machine learning models automatically analyze and update inventory data without human intervention. The system serves itself by continuously capturing images and processing them to maintain accurate inventory records across all storage areas.
3Reliability
If workers are assigned to manually inspect product storage areas, then inventory status is monitored, but workers are unavailable for other tasks
Solution Approach 1:
The patent replaces human workers with an automated image capture and processing system for inventory monitoring. Image capture devices continuously or periodically photograph storage areas, and machine learning models automatically process these images to detect stock status changes, ensuring reliable monitoring while freeing workers for value-added tasks.
Solution Approach 2:
The automated system enables continuous inventory monitoring without the interruptions inherent in manual inspection. Image capture devices can continuously capture images of storage areas, and the machine learning models continuously process these images to detect changes in real-time, providing uninterrupted inventory visibility.
Data Source
AI summary
In some embodiments, apparatuses and methods are provided herein useful to processing captured images. In some embodiments, there is provided a system for processing captured images of objects at a product storage facility including a trained machine learning model stored in a memory; and a control circuit. The control circuit may obtain an image at the product storage facility; cluster objects depicted in the image that have same product identifiers into a corresponding group; determine coordinates of each bounding box of each clustered object in the corresponding group; determine a bounding box representative depth value of pixels inside the bounding box of each clustered object; determine an overall representative depth value of the corresponding group based on bounding box representative depth values of clustered objects; and exclude the clustered objects from identified objects in the image upon a determination that the overall representative depth value is greater than a threshold.