Image Processing Annotation Merging by Relevance Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing systems face challenges in accurately and efficiently integrating annotations recorded by different devices, leading to inconsistent and redundant annotations in image files.

Innovation Solution

An image processing apparatus and method that integrates and displays annotations from multiple devices by determining relevance and integrating overlapping or related annotations, using a neural network for subject detection and metadata management.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If multiple recording devices record annotations in one image file, then the annotation coverage and detection capabilities are improved, but the annotation consistency and redundancy increase

Engineering Contradiction:
Improveannotation coverageVSAvoidannotation consistency
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary actions by establishing annotation integration rules and device identification mechanisms before multiple devices record annotations. This includes setting up the framework for determining device identities, comparing annotation regions, and applying integration strategies to prevent redundancy before it occurs

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements feedback mechanisms by continuously monitoring annotations from multiple devices, comparing them against existing annotations, and adjusting the integration process based on overlap detection and device reliability assessment. This feedback loop ensures consistent annotation quality while maintaining versatility

Inventive Principle:
Principle #23Feedback

2Loss of substance

If annotations from multiple devices are integrated, then the redundancy is reduced, but the processing complexity increases

Engineering Contradiction:
Improveannotation redundancyVSAvoidprocessing complexity
Core Design Contradiction:
Loss of substanceVSDevice complexity

Solution Approach 1:

The system extracts and separates the annotation integration function from the overall image processing workflow. By isolating the comparison and integration logic into a dedicated module, it reduces redundancy efficiently while managing processing complexity through functional decomposition

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system changes parameters by adjusting the integration threshold and device identification criteria based on the specific application context. This allows flexible control over the balance between redundancy reduction and processing complexity without requiring complete redesign of the integration architecture

Inventive Principle:
Principle #35Parameter changes

3Quantity of substance

If annotation integration is performed, then the storage efficiency is improved, but the processing time increases

Engineering Contradiction:
Improvestorage efficiencyVSAvoidprocessing time
Core Design Contradiction:
Quantity of substanceVSLoss of time

Solution Approach 1:

The system performs preliminary actions by pre-establishing integration rules and device profiles before annotation processing. This preparation work enables faster real-time integration by avoiding repeated analysis of device characteristics and integration strategies during the actual annotation merging process

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system applies discarding and recovering by identifying and discarding redundant annotations through comparison algorithms, then recovering only the unique and valuable annotation data. This selective process reduces storage requirements while minimizing processing time by avoiding unnecessary analysis of duplicate information

Inventive Principle:
Principle #34Discarding and recovering

Data Source

PatentEP4614351A1Image processing apparatus, method for controlling image processing apparatus, computer-readable storage medium, and computer program
Publication Date: 2025.09.10 CANON KK
  • EP4614351A1 patent drawingFigure 1
  • EP4614351A1 patent drawingFigure 2
  • EP4614351A1 patent drawingFigure 3

AI summary

An image processing apparatus includes acquisition means (201, 300) configured to acquire an image file including a plurality of annotations, determination means (201, 1000) configured to determine a relevance between the plurality of annotations, and control means (201, 1002) configured to perform control to integrate annotations determined to be relevant by the determination means among the plurality of annotations into an integrated annotation and output the integrated annotation.