Vending Machine Recognition via Peripheral Image Region Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for automatically collecting merchandise information from vending machines, such as prices, are limited as they can only read character information from the merchandise surface and require pre-registered attribute information, making it difficult to collect prices from unmanaged vending machines.
Innovation Solution
A vending machine recognition apparatus that receives images of vending machines, recognizes merchandise, and identifies prices from peripheral image regions, allowing for the collection of merchandise information including prices from unmanaged vending machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If character information is merely read from the merchandise surface using existing techniques, then the reading process is simple, but it becomes impossible to automatically collect price information when prices are not written on the merchandise itself
Solution Approach 1:
The patent transitions from reading only the merchandise surface (2D character recognition) to analyzing the peripheral image region surrounding the merchandise (spatial dimension expansion). This allows the system to capture price information displayed on vending machine panels or labels adjacent to the merchandise, not just characters directly on the product surface.
Solution Approach 2:
The patent introduces a peripheral image region as an intermediary zone between the merchandise and the price information. This peripheral region contains contextual information (prices, descriptions) that mediates the connection between the merchandise image and its pricing data, enabling automatic collection without direct character reading from the merchandise itself.
2Productivity
If pre-registered attribute information is used for collation, then the collation process is efficient, but it cannot collect prices from unmanaged vending machines without prior registration
Solution Approach 1:
The patent enables unmanaged vending machines to automatically provide their own price information through image capture and peripheral region analysis. The system extracts price data directly from the vending machine's own display or labeling without requiring external registration or pre-existing database entries, allowing self-service information collection.
Solution Approach 2:
The patent performs preliminary extraction of price information from the peripheral image region during the image capture phase. By pre-identifying and extracting price data from the captured image before collation, the system prepares information that can be directly used for merchandise identification and pricing, eliminating the need for subsequent manual registration.
3Measurement precision
If manual collection of merchandise information is performed, then information accuracy is high, but work cost and work time increase significantly
Solution Approach 1:
The patent replaces the mechanical manual collection process with an automated image processing system. The recognition unit automatically extracts merchandise images and peripheral price information, substituting human visual inspection and data entry with computational image analysis, thereby maintaining accuracy while dramatically reducing time and labor costs.
Data Source
AI summary
A vending machine recognition apparatus includes: a receiving unit configured to receive a captured image obtained by capturing an image of a vending machine; and a recognition unit configured to recognize merchandise from the captured image and recognize a price of the merchandise from a peripheral image region of an image region of the recognized merchandise.