ROI-Centric Image Generation with Machine Learning Rescaling

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing electronic devices rely heavily on external tools and user skill/creativity for editing and resizing objects in images, leading to inefficient and often unnatural results, especially when capturing images with specific perspectives or object arrangements.

Innovation Solution

An electronic device uses a Machine Learning model to identify and rescale Regions of Interest (ROIs) based on object importance and redundancy scores, generating an ROI centric image that includes rescaled first and second ROIs and summarized non-ROIs, enhancing aesthetic value and user experience.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If existing external tools and user skill are used for editing and resizing objects in images, then some level of image modification is achieved, but the process is time-consuming and results are often unnatural

Engineering Contradiction:

Improveimage editing efficiencyVSAvoidtime required for manual editing

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs automatic image editing and object resizing without requiring user intervention or external tools. The machine learning model autonomously identifies objects, determines their importance, and generates edited images with natural appearances, enabling the system to serve itself rather than requiring user expertise or external software assistance

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical editing processes with an automated machine learning-based system. Instead of requiring users to manually adjust objects in external tools like Photoshop, the system uses neural networks to automatically detect, select, and resize objects while maintaining natural appearances, substituting human expertise with computational intelligence

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Ease of operation

If manual editing and resizing of objects is performed, then specific image modifications can be achieved, but it requires user expertise and creativity that not everyone possesses

Engineering Contradiction:

Improveuser-friendliness of image editingVSAvoiddependency on external tools and user skill

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The system autonomously performs image editing tasks without requiring user expertise. The machine learning model automatically identifies objects of interest, determines their importance scores, and generates edited images with natural appearances, making the system self-sufficient and eliminating the need for users to possess specialized editing skills or use complex external tools

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent introduces a machine learning intermediary layer between the user and the image editing process. Instead of users directly manipulating complex editing tools, the system uses an AI intermediary to automatically detect objects, understand their spatial relationships, and generate natural-looking edits, simplifying the interface while maintaining sophisticated editing capabilities

Inventive Principle:

Principle #24Intermediary (Mediator)

3Manufacturing precision

If objects are resized and repositioned in 3D planar space, then desired perspective changes can be achieved, but the process is time-consuming and results may appear unnatural

Engineering Contradiction:

Improveperspective accuracy of edited imageVSAvoidtime required for perspective adjustment

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical perspective adjustment with an automated machine learning system. The neural network automatically analyzes the original image, identifies objects and their spatial relationships, and generates edited images with corrected perspectives in a single operation, eliminating the need for time-consuming manual manipulation while achieving accurate and natural results

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12374074B2Method and electronic device for automatically generating region of interest centric image

Publication Date: 2025.07.29 SAMSUNG ELECTRONICS CO LTD

AI summary

A method for automatically generating a Region Of Interest (ROI) centric image in an electronic device is provided. The method includes receiving an image frame(s), where the image frame(s) includes a plurality of objects. Further, the method includes identifying a first ROI, a second ROI, and a non-ROI in the image frame(s). Further, the method includes rescaling the second ROI in the image frame(s), summarizing the non-ROI in the image frame(s), and automatically generating the ROI centric image, where the ROI centric image includes the rescaled-first ROI, the rescaled-second ROI, the rescaled-non-ROI, and the summarized non-ROI.