Encoded Substrate Grids for Occlusion-Free Shelf Inventory
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the physical retail industry, existing inventory methods rely on manual labor to count products on shelves, leading to inefficiencies and high labor costs, as object recognition technologies struggle to accurately count products due to occlusion issues when goods are closely arranged, allowing only the frontmost products to be identified.
Innovation Solution
An encoded substrate with a two-dimensional array of grids, each containing unique first and second patterns, is used to generate images that can be processed by a camera and processor to output coordinates, enabling accurate counting of products by identifying uncovered patterns, thus overcoming occlusion challenges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If goods are arranged closely on the shelf to effectively use space, then space utilization is improved, but the frontmost product occludes the inner products, making them unidentifiable by camera devices
Solution Approach 1:
The invention divides the identification task into two independent components: the encoded substrate (first pattern) that remains visible and identifies the product category, and the product image (second pattern) that identifies the specific product. This segmentation allows the frontmost product's encoded substrate to represent both itself and the occluded inner products, solving the occlusion problem while maintaining close arrangement.
Solution Approach 2:
The encoded substrate acts as an intermediary element that carries identification information for the entire product category. Instead of requiring direct visual access to each product, the encoded substrate on the frontmost product serves as a mediator that provides information about the occluded inner products through its first pattern.
2Measurement precision
If manual labor is used to count products on shelves, then product identification accuracy is improved, but labor costs and time consumption increase significantly
Solution Approach 1:
The invention replaces the mechanical manual counting process with an automated image recognition system. The camera device captures images of the encoded substrate, and the processor automatically identifies and counts products by recognizing the first and second patterns, eliminating the need for manual labor while maintaining high accuracy.
Solution Approach 2:
The encoded substrate is designed to be self-identifying through its unique first and second patterns. The system enables automatic product identification and counting without human intervention, as the encoded substrate on each product (or the frontmost product representing the category) provides all necessary information for the processor to determine product type and quantity.
3Speed
If only the frontmost product is identified by camera devices, then identification speed is improved, but the actual quantity of products cannot be calculated
Solution Approach 1:
The encoded substrate serves multiple functions simultaneously: it identifies the product category through the first pattern, identifies the specific product through the second pattern, and enables quantity calculation by being placed on each product (or the frontmost product representing the category). This multi-functionality allows a single identification action to provide comprehensive information for inventory management.
Data Source
AI summary
An encoded substrate to be filmed by a camera device for generating an image is provided. The encoded substrate includes a plurality of grids arranged in a form of two-dimensional array, wherein each grid includes a first pattern and a second pattern not overlapped with each other. The first pattern corresponds to a first-dimensional encoded value, and the second pattern corresponds to a second-dimensional encoded value. The image is processed by a processor for scanning the plurality of grids. In a first-dimensional direction, the processor outputs a first coordinate according to at least two first patterns corresponding to at least two consecutive grids among the plurality of grids. In a second-dimensional direction, the processor outputs a second coordinate according to at least two second patterns corresponding to at least two consecutive grids among the plurality of grids.