Automated Shelf Image Analysis for Planogram Compliance Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems are inefficient in determining and correcting the physical configuration of products on store shelves, especially in large retail environments, leading to time-consuming and error-prone processes for assessing and re-stocking items.
Innovation Solution
A system comprising a cloud-based application server, capture devices, and a database that acquires and processes images of store shelves, generates corrective actions based on planogram reference data, and provides intuitive interfaces for users to address location errors, using algorithms to efficiently identify product locations and determine necessary actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual product surveys and corrections are performed by product representatives, then product positions can be assessed and corrected, but the process becomes extremely time-consuming and error-prone in large retail environments
Solution Approach 1:
The patent replaces manual mechanical inspection by product representatives with an automated image recognition system using capture devices (cameras) and processing algorithms. The system captures images of shelves, automatically identifies product locations and configurations, and generates corrective actions, eliminating the time-consuming manual survey process while maintaining or improving accuracy.
Solution Approach 2:
The system enables self-service by allowing the automated image recognition system to independently assess product configurations, identify discrepancies from planograms, and generate corrective action recommendations without requiring human intervention for the assessment process. This automates what previously required manual human labor.
2Productivity
If manual methods are used to assess and correct product configurations in long store aisles with hundreds of items, then product positions can be evaluated, but the process becomes ad-hoc and inefficient
Solution Approach 1:
The system replaces inefficient manual assessment methods with automated image capture and processing. Capture devices quickly photograph entire shelf sections, and algorithms automatically analyze the images to identify product placement issues, generating corrective actions efficiently without the ad-hoc nature of manual processes.
Solution Approach 2:
The patent introduces an intermediary processing system that acts as a mediator between the physical product configuration and the corrective actions. The image recognition system and algorithm processor serve as intermediaries that translate visual shelf data into structured corrective action recommendations, simplifying the operation for users.
3Measurement precision
If conventional systems are used to determine and guide correction of errors in physical configuration, then some product placement issues can be identified, but the system fails to provide an efficient solution for large-scale retail environments
Solution Approach 1:
The patent segments the complex task of product configuration management into distinct functional modules: image capture by portable devices, image processing and analysis by algorithms, discrepancy identification by comparing against planograms, and corrective action generation. This segmentation handles large-scale retail environments efficiently while maintaining precision through specialized processing at each stage.
Solution Approach 2:
The system uses planogram data as an intermediary reference standard to objectively measure and identify product placement discrepancies. The planograms serve as a mediator between the desired configuration and the actual observed configuration, enabling precise measurement of deviations without requiring complex judgment calls.
Data Source
AI summary
A system includes acquisition of an electronic image of items disposed in first locations, determination, based on the electronic image, of a visit image result comprising a first one or more linear groups, a first one or more facings per linear group and a first zero or more product identifiers per facing, determination of planogram reference data comprising a second one or more linear groups, a second one or more facings per linear group and a second zero or more product identifiers per facing, and determination of one or more corrective actions based on the visit image result and on the planogram reference data.