Automated Image Clustering for Product Inventory Labeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual inventory inspection in large product storage facilities is time-consuming and costly, as workers need to visually check hundreds of shelves and thousands of products, diverting resources from other tasks.
Innovation Solution
A system that uses image capture devices to capture images of product storage areas, which are then processed by a machine learning model to label objects, allowing for automated clustering and labeling of products, reducing the need for manual inspection.
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 time-consuming and increases operational costs
Solution Approach 1:
The patent replaces the mechanical manual inspection system with an automated image-based detection system. Image capture devices photograph product shelves, and machine learning models automatically analyze the images to identify stocked and out-of-stock products, eliminating the need for workers to physically inspect each product while maintaining high accuracy
Solution Approach 2:
The system creates visual copies (images) of the product storage areas and processes these copies through machine learning models. This allows the system to analyze product inventory status without physically interacting with the products or requiring workers to be present, thereby reducing inspection time while maintaining measurement precision
2Area of stationary object
If manual inspection is deployed across large product storage facilities, then all product areas can be covered, but operational costs significantly increase
Solution Approach 1:
The patent replaces human workers with automated image capture devices and machine learning processing. This substitution allows the system to cover large product storage facilities without the escalating operational costs associated with deploying more manual laborers, as the automated system incurs only fixed technological costs regardless of facility size
Solution Approach 2:
The machine learning model serves multiple functions: it identifies out-of-stock products, detects product locations, and can potentially recognize various product types. This multi-functionality allows a single automated system to handle diverse inventory management tasks across the entire facility, reducing the need for specialized manual inspection teams for different product categories
3Loss of information
If workers are assigned to manual inventory inspection, then product stock status can be determined, but other important tasks cannot be performed
Solution Approach 1:
The system creates digital copies of inventory data through image capture and processing. These digital records provide complete inventory status information without requiring worker presence, allowing workers to be reassigned to value-added tasks while the automated system continuously monitors and updates inventory information
Solution Approach 2:
The inventory management system becomes self-service through automated image capture and machine learning analysis. The system independently performs inventory status determination without human intervention, freeing workers from this repetitive task and enabling them to focus on higher-productivity activities such as customer service, product arrangement, and exception handling
Data Source
AI summary
In some embodiments, apparatuses and methods are provided herein useful to labeling objects in captured images. In some embodiments, there is provided a system for labeling objects in images captured at a product storage facility including a control circuit and a user interface. The control circuit is configured to select a set of unprocessed images; receive a selected configuration based on data resulting from iteratively processing the set of unprocessed images; cluster each unprocessed image into a corresponding group based on the selected configuration; select a plurality of clustered images from each of the plurality of groups; and output the plurality of clustered images from each group. The user interface is configured to: display each clustered image; and receive a user input labeling one or more objects shown in each clustered image resulting in a labeled dataset used to train a machine learning model.