Face Image Classification for Mobile Terminal Photo Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Recent cellular phone terminals face the challenge of managing and classifying a growing number of image files, as users must manually update and associate new, changed, or deleted image files with corresponding content data, leading to a cumbersome and often incomplete classification process.
Innovation Solution
An information terminal equipped with an image detector/register, a face image detector, a similar face classification section, and a similar face display section, which detects unregistered or updated image files, extracts and classifies face images by comparing features, and displays grouped images on the screen, facilitating automatic classification and association with other content data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual classification of image files is performed, then image files can be associated with content data, but user burden increases and classification completeness decreases
Solution Approach 1:
The system automatically detects unregistered or updated image files, extracts face images, compares them with existing face images using feature amounts, and performs classification without user intervention. This self-service mechanism resolves the contradiction by eliminating manual classification burden while maintaining high classification completeness through automated feature-based matching.
Solution Approach 2:
The patent replaces the mechanical manual classification process with an automated computer-based system that uses image processing and feature comparison algorithms. The face image detection, feature extraction, and similarity comparison are performed automatically by the information processing unit, substituting human manual operations with automated computational processes.
2Productivity
If automated face image classification is implemented, then classification efficiency improves, but system complexity increases
Solution Approach 1:
The classification system is divided into distinct functional modules: an image file detection unit that identifies unregistered or updated files, a face image detection unit that extracts faces from images, a feature comparison unit that calculates similarity, and a classification unit that assigns categories. This segmentation allows each module to perform a specific function efficiently, improving overall classification productivity while managing system complexity through modular design.
Solution Approach 2:
The patent introduces a face image as an intermediary element that mediates between the raw image files and the classification result. By extracting and comparing face images as intermediate representations, the system achieves efficient automated classification without requiring direct complex analysis of entire image files, thus balancing productivity improvement with acceptable system complexity.
3Measurement precision
If all image files are manually checked and classified, then classification accuracy improves, but time consumption increases
Solution Approach 1:
The system performs preliminary detection of unregistered or updated image files and preliminary extraction of face images before the actual classification process. By preparing these intermediate results in advance, the system reduces the time required for the main classification operation while maintaining accuracy through subsequent feature-based comparison with existing classified images.
Solution Approach 2:
The patent replaces time-consuming manual checking and classification with automated image processing and feature comparison. The computer-based system rapidly analyzes image files, extracts relevant features, and performs classification algorithmically, achieving both high classification accuracy and reduced time consumption compared to manual processes.
Data Source
AI summary
To make it possible to significantly reduce a burden on a user at a time of a classification of image files added or updated, for example, and thus facilitate association with other content data. An image detector/register (20) detects an image file that is unregistered or updated, and structures and manages a list of the image files. A face rectangular detector (21) detects a face rectangle in the image files detected, associates information on the face rectangle with the image file, and stores the information. A similar face classification section (22) calculates an amount of feature of the face rectangle, compares the amount of feature calculated with amounts of feature of different face rectangles already calculated and classified into groups, calculates a similarity between the face rectangle and the face rectangles of the different face rectangles in the respective groups, and classifies and manages the face rectangles in accordance with a result of the similarity calculation. A similar face display section (23) displays at least one image file including the face rectangle classified into the same group on a display screen.