Dual Segmentation Image Analysis for Coating Region Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image analysis methods struggle to accurately identify objects with both a base material and a coating region, as instance segmentation and semantic segmentation have limitations in distinguishing and defining contours and coating regions.

Innovation Solution

An image analysis system employing both instance and semantic segmentation models to set masks for objects, combining these to accurately identify contours and coating regions, using a computational model for image segmentation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If deep learning-based image recognition is used to achieve high precision in identifying objects of interest, then measurement precision is improved, but device complexity and processing time increase

Engineering Contradiction:
Improveidentification precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The image analysis process is segmented into multiple stages: first identifying candidate regions using simplified criteria, then applying deep learning-based recognition only to these candidates. This segmentation reduces the overall system complexity while maintaining high identification precision for objects of interest.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary filtering to identify candidate regions before applying complex deep learning recognition. By pre-processing the image to locate potential objects of interest, the system reduces the computational burden on the deep learning model and improves overall processing efficiency.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If deep learning-based image recognition is used to achieve high precision in identifying objects of interest, then measurement precision is improved, but processing time increases

Engineering Contradiction:
Improveidentification precisionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The image analysis is divided into fast candidate region detection and subsequent precise deep learning recognition. This segmentation allows the system to quickly narrow down search areas, reducing the time required for comprehensive image analysis while maintaining high precision.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Candidate region identification is performed as a preliminary step before applying time-consuming deep learning recognition. This preliminary action significantly reduces the amount of data that requires intensive processing, thereby decreasing overall processing time.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If comprehensive image analysis is performed to ensure no objects of interest are missed, then reliability is improved, but productivity decreases

Engineering Contradiction:
Improvedetection reliabilityVSAvoidanalysis speed
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The analysis process is segmented into candidate region identification followed by focused deep learning recognition. This approach maintains detection reliability by ensuring thorough examination of potential objects while improving productivity through efficient resource allocation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

By performing preliminary candidate region identification, the system prepares the data in advance for rapid deep learning recognition. This preliminary action ensures comprehensive coverage of potential objects while optimizing the subsequent recognition phase for speed.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4618014A1Image analysis system, image analysis method, and image analysis program
Publication Date: 2025.09.17 RESONAC CORP
  • EP4618014A1 patent drawingFigure 1
  • EP4618014A1 patent drawingFigure 2
  • EP4618014A1 patent drawingFigure 3

AI summary

An image analysis system acquires a target image showing one or more objects each of which has a base material and a coating region on the base material, inputs the target image to an instance segmentation model executing instance segmentation to set at least one instance mask corresponding to at least one of the one or more objects, inputs the target image to a semantic segmentation model executing semantic segmentation to set a semantic mask indicating a contour and a coating region of an object set consisting of the at least one object, and identifies, for each of the at least one object, the contour and the coating region of the object based on the at least one instance mask and the semantic mask.