Image Processing Apparatus Inversion for Unexpected Content Discovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing image processing technologies fail to automatically extract enjoyable images from large collections, often missing images that are not initially recognized as important or valuable to the user.

Innovation Solution

An image processing apparatus and method that includes detection devices to identify remaining images from two groups, outputting images with attribute frequencies below a threshold, ensuring the output image is enjoyable and unexpected for the user, utilizing a smartphone with integrated detection and processing capabilities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If image processing extracts only images recognized as important or valuable to the user, then the extraction accuracy for user-important images is improved, but enjoyable images that are not recognized as important are lost

Engineering Contradiction:

Improveextraction accuracyVSAvoidloss of enjoyable images

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent inverts the conventional approach by not directly extracting enjoyable images, but rather extracting images that are NOT recognized as important or valuable. By selecting from the complement set (remaining images after removing important/valuable ones), the system discovers enjoyable images that were previously overlooked, thus resolving the contradiction between extraction accuracy and information loss.

Inventive Principle:

Principle #13The other way round (Inversion)

2Adaptability or versatility

If conventional image processing focuses on extracting important or valuable images, then the relevance to user needs is improved, but the diversity and unexpectedness of output images deteriorates

Engineering Contradiction:

Improverelevance to user needsVSAvoiddiversity of output

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent applies inversion by selecting images from the opposite category - instead of choosing images that match user needs (important/valuable), it chooses images that do not match these criteria. This inverted selection strategy inherently increases diversity and unexpectedness while maintaining a connection to user preferences through the frequency-based filtering mechanism.

Inventive Principle:

Principle #13The other way round (Inversion)

Solution Approach 2:

The patent changes the selection parameter from high frequency (important/valuable images) to low frequency (unexpected images). By setting a threshold on attribute frequency and selecting images below this threshold, the system transforms the selection criterion to achieve both relevance and diversity, resolving the contradiction between adaptability and output complexity.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS11200651B2Image processing apparatus, image processing method, and image processing program

Publication Date: 2021.12.14 FUJIFILM CORP

AI summary

Provided are an image processing apparatus, an image processing method, and an image processing program capable of obtaining an enjoyable image for a user unexpectedly. The user selects images from image groups imaged from January to December 2017 as an image to be printed. From not-selected remaining images, an image having an attribute in which a frequency for the attribute contained in each image is equal to or less than a threshold is found and printed. Since the obtained image is an image irrelevant to the images selected by the user, the image unexpected for the user can be obtained.