Image Evaluation Using Representative Impression Values

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image evaluation methods fail to accurately assess images based on user-preferred impressions and objective quality, as they rely solely on visual features or specific attributes like face position, leading to incomplete evaluation of image impressions and qualities.

Innovation Solution

An image evaluation apparatus and method that selects a representative image from a group, calculating an evaluation value based on the difference between the impression value of the selected image and each input image, considering both impression value and image quality, to provide a comprehensive evaluation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If visual features or specific attributes (face position, brightness, shake amount) are used for image evaluation, then the evaluation process becomes automated and efficient, but the evaluation cannot capture user-preferred impressions and objective quality simultaneously

Engineering Contradiction:
Improveevaluation efficiencyVSAvoidevaluation accuracy
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent merges two separate evaluation dimensions: impression value (subjective user preference) and image quality (objective quality metrics). By combining these previously separate evaluation criteria into a unified image evaluation value, the system achieves both automated processing and comprehensive assessment that captures both user preference and objective quality.

Inventive Principle:
Principle #5Merging (Combining)

2Reliability

If only visual features are used for image evaluation, then the evaluation can be performed objectively, but images that cannot be expressed from visual features cannot be evaluated

Engineering Contradiction:
Improveevaluation objectivityVSAvoidevaluation coverage
Core Design Contradiction:
ReliabilityVSAdaptability or versatility

Solution Approach 1:

The patent introduces a new evaluation parameter dimension by distinguishing between impression value (subjective) and image quality (objective). This parameter change allows the system to evaluate both visual features and non-visual characteristics, expanding evaluation coverage while maintaining objectivity through the separate image quality assessment.

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If a subject image is determined only by face position or similar specific attributes, then the selection process is simplified, but an image having a desired impression cannot be highly evaluated

Engineering Contradiction:
Improveselection simplicityVSAvoidimpression evaluation accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent segments the image evaluation into two distinct components: impression value evaluation (capturing user preference) and image quality evaluation (capturing objective quality). This segmentation allows the system to maintain simple automated processing while accurately evaluating both impression and quality, resolving the contradiction between simplicity and accuracy.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS11210771B2Image evaluation apparatus, image evaluation method, and image evaluation program using impression values of representative images
Publication Date: 2021.12.28 FUJIFILM CORP
  • US11210771B2 patent drawing
  • US11210771B2 patent drawing
  • US11210771B2 patent drawing

AI summary

There are provided an image evaluation apparatus, an image evaluation method, and an image evaluation program capable of highly evaluating an image having an impression of a user's preference. A plurality of images included in a first image group are input to an image evaluation apparatus (step 21). A plurality of representative image candidates are displayed (step 22), and the user selects a desired representative image from the representative image candidates (step 23). The difference between the impression value of the representative image and the impression value of the input image is calculated (step 24), and the image quality of the input image is determined (step 25). An image evaluation value is calculated so as to become higher as the difference between the impression values becomes smaller and the image quality becomes higher (step 26).