Image Grouping via Inertial Profile Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In asset management, large collections of images from various sources are often unstructured and difficult to filter effectively, leading to irrelevant results due to lack of distinguishing features and reliance on incomplete metadata, especially when images are nondescript or captured for multiple purposes.

Innovation Solution

A method that groups images based on inertial profiles derived from acceleration data and camera settings, including focus distance, focal length, exposure, and brightness values, to differentiate between image capture activities and associate them with specific tasks, using a processor and memory device to normalize and form image groups.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If manual annotation is used to annotate images during capture or transfer, then image data can be distinguished and organized, but labor costs and time consumption increase significantly

Engineering Contradiction:

Improveimage distinction informationVSAvoidannotation time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The system performs preliminary action by capturing inertial data (acceleration, orientation) at the moment of image capture, storing it alongside the image metadata. This pre-captured motion information is later used to automatically group images, eliminating the need for manual annotation while preserving image distinction information.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces the mechanical/manual system of human annotation with an automated computational system that uses inertial sensor data and image processing algorithms to automatically group and distinguish images based on capture context, significantly reducing labor time while maintaining information quality.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Ease of manufacture

If location metadata is used to cluster images into groups, then images can be organized automatically, but irrelevant images are included in query results because different capture purposes are not distinguished

Engineering Contradiction:

Improveautomatic image organizationVSAvoidquery result relevance

Core Design Contradiction:

Ease of manufactureVSReliability

Solution Approach 1:

The system adds another dimension to image organization by incorporating inertial profile data (motion patterns, orientation, acceleration) alongside location metadata. This multi-dimensional approach enables differentiation of capture purposes - for example, distinguishing between images taken while walking versus standing still - thereby improving query result relevance while maintaining automatic organization.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Productivity

If automatic recognition and tagging of image content is implemented, then manual labor is reduced, but nondescript images lacking distinguishing features produce unuseful groupings for asset management

Engineering Contradiction:

Improveimage processing efficiencyVSAvoidimage distinction capability

Core Design Contradiction:

ProductivityVSLoss of information

Solution Approach 1:

The patent introduces an intermediary - the inertial profile - that captures contextual information about the capture event itself (motion, orientation, stability) rather than relying solely on visual content analysis. This intermediary provides distinguishing features for nondescript images, enabling meaningful groupings while maintaining high processing efficiency through automated inertial data capture and comparison.

Inventive Principle:

Principle #24Intermediary (Mediator)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

This approach enhances the utility of image data by accurately grouping images based on their capture context, reducing manual annotation costs and improving the relevance of search results without increasing data collection complexity.

Implementation Method 1

determining an inertial profile for the plurality of images based on acceleration data of the image capture device

Methodology Applied

Scientific EffectAcceleration: Accelerometer

Data Source

PatentUS11138258B2System and method of grouping images

Publication Date: 2021.10.05 CANON KK

AI summary

A system and method of grouping images captured using an image capture device. The method comprises receiving a plurality of images, each of the plurality of images having associated camera settings; and determining an inertial profile for the plurality of images based on acceleration data of the image capture device and an imaging entity at pre-determined length of time before and after capture of the each of the plurality of images. The method further comprises forming image groups from the received plurality of images based on the determined inertial profile, and the associated camera settings.