Product Identification via Container and Product Feature Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to accurately identify groups of products that appear similar to each other through image analysis due to similarities in shape, color, and size.
Innovation Solution
A processing apparatus and method that computes identification scores for containers and products based on image features, using machine learning models and template matching techniques to determine and estimate identification information, allowing for accurate differentiation between similar products.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If image analysis is used to identify products, then identification speed is improved, but identification accuracy deteriorates for similar products
Solution Approach 1:
The identification process is segmented into two independent analysis streams: container identification and product identification. The container identification score computation means analyzes container features (shape, color, size) separately, while the product identification score computation means analyzes product features separately. This segmentation allows each stream to specialize in detecting specific features, improving overall accuracy for similar products while maintaining high processing speed through parallel computation.
2Device complexity
If only product features are analyzed for identification, then processing complexity is reduced, but identification accuracy deteriorates for similar products
Solution Approach 1:
The system merges container feature analysis and product feature analysis into a unified identification framework. The container identification score and product identification score are computed separately but combined in the first identification means to make the final identification decision. This merging allows the system to leverage both container and product features without significantly increasing processing complexity, as each feature type is analyzed independently and then integrated.
3Measurement precision
If multiple container types are used for different products, then product differentiation is improved, but device complexity and cost increase
Solution Approach 1:
The system extracts and analyzes container features as a separate identifiable characteristic independent of product features. By taking out container identification as a distinct analysis stream, the system can differentiate products based on their containers without requiring multiple container types for each product variant. The container identification score computation means extracts container-specific features (shape, color, size) that serve as additional identification criteria, reducing the need for container diversity while maintaining accurate product differentiation.
Data Source
AI summary
This invention provides a processing apparatus (10) including a container identification score computation unit (11) that computes an identification score of a container for a product, based on an image to be processed containing the product, a product identification score computation unit (12) that computes an identification score of the product, based on the image to be processed, and a first identification unit (13) that decides identification information of the product, based on the computed identification score of the container and the computed identification score of the product.