Automatic Image Tagging via Event-Based Classification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Users face difficulties in classifying and managing large numbers of digital images captured by devices, as existing methods fail to automatically categorize images by events or contexts beyond chronological order, leading to inefficient search processes.

Innovation Solution

An image apparatus and method that automatically adds tag information to photographs based on recognized shapes or pre-stored images, allowing for easy classification and storage of images by using a photographing unit, storage unit, and controlling unit to register and display tag images alongside the photographed images.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If images are classified chronologically by date, then images are organized by time, but images photographed on the same date cannot be classified into different groups according to events

Engineering Contradiction:
ImproveImage classificationVSAvoidEvent-based categorization
Core Design Contradiction:
Ease of operationVSAdaptability or versatility

Solution Approach 1:

The patent segments images into multiple classification dimensions: chronological order (date) and event-based categories (location, activity, object). This allows images from the same date to be further divided into different event groups, resolving the contradiction between simple time-based organization and detailed event-based categorization.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds new classification dimensions beyond time by introducing event-based categories such as location, activity, and object type. This dimensional expansion enables images to be organized both by when they were taken and by what they depict, achieving both chronological simplicity and event-specific detail.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If a user personally checks all images to search for a specific image, then all images can be reviewed, but very long time is wasted

Engineering Contradiction:
ImproveImage search accuracyVSAvoidSearch time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent performs preliminary classification of images into event-based categories and locations at the time of capture. This pre-organization allows users to quickly navigate to specific event groups or locations without reviewing all images chronologically, significantly reducing search time while maintaining accurate retrieval.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces event metadata and location information as intermediary elements between the user's search query and the actual images. Users can search through these intermediary categories (e.g., 'beach', 'swimming', '2024-01-15') to quickly locate specific images without manually checking each one.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Extent of automation

If automatic tag information is added to images, then images can be automatically classified, but device complexity increases

Engineering Contradiction:
ImproveImage taggingVSAvoidSystem structure
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The patent implements self-service automation where the imaging device automatically extracts metadata (location, time, event type) and attaches relevant tags to images during the capture process. This eliminates the need for manual tagging while using relatively simple automated extraction techniques, balancing automation extent with system complexity.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS8704914B2Apparatus to automatically tag image and method thereof
Publication Date: 2014.04.22 SAMSUNG ELECTRONICS CO LTD
  • US8704914B2 patent drawing
  • US8704914B2 patent drawing
  • US8704914B2 patent drawing

AI summary

An apparatus to automatically tag an image and a method thereof conveniently classifies images by acquiring information of a tag image when an image is photographed and stored, automatically adding tag information to the photographed image based on the information of a tag image, generating an image file, and storing the image file. The apparatus and method acquire an image and record a photographed image based on classification information of an image which is automatically tagged when the image is acquired and stored.