Person Identification Using Relationship Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image classification devices struggle to accurately retrieve images of the same person when their facial appearance changes over time, such as from childhood to adulthood, due to differences in feature points, leading to misclassification.
Innovation Solution
An information processing apparatus and method that detects individuals in images, determines relationships between people, and uses person relationship information to specify individuals even with changed facial appearances, allowing accurate retrieval and printing of images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image classification is performed based on face image feature points, then individual people can be specified and images can be classified, but when a person's facial appearance changes (e.g., from childhood to adulthood), the same person cannot be accurately retrieved
Solution Approach 1:
The patent transitions from two-dimensional face image feature point comparison to incorporating temporal dimension by analyzing multiple images across different time periods. The system extracts feature points from multiple images taken at different times and determines whether the same person appears in all images, thereby adding the time dimension to person specification and enabling accurate retrieval despite facial appearance changes.
2Ease of operation
If multiple copies of images are printed for each detected person, then each person's images can be printed individually, but this causes wastage when multiple people are in the same image
Solution Approach 1:
The patent merges the printing operations for multiple people detected in the same image. By determining that multiple people are present in a single image, the system consolidates the printing process into one operation rather than printing separate copies for each person, thereby reducing printed image waste while still enabling individual printing capability when needed.
Data Source
AI summary
An information processing apparatus includes: a storage unit which stores person relationship information representing relationships between people as a subject in a storage medium; an acquisition unit which acquires image data generated by imaging people as a subject; a detection unit which detects each person in an image; a specification unit which specifies each person detected from the image by the detection unit; and a determination unit which determines the relationship between multiple people detected from the image. When at least one person from among the people detected from the image is specified and another person is unable to be specified, the specification unit may specify the at least one other person on the basis of the relationship between the multiple people and the person relationship information.