Endoscope Mask Boundary Detection via Two-Stage Shape Estimation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Endoscope mask detection is often inaccurate due to frequent shifts in the mask position, making it difficult to distinguish between image-providing and non-image-providing regions, especially in medical imaging where real-time processing is crucial.

Innovation Solution

An image processing device with edge detection, circle, and ellipse estimation units that use luminance values and standard deviation to accurately detect and correct the mask boundary, employing two-stage estimation to improve detection accuracy and reduce erroneous edge points.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a single-stage mask detection method is used, then the device complexity is reduced, but the measurement precision of the mask boundary deteriorates

Engineering Contradiction:

Improvedetection method complexityVSAvoidmask boundary detection accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The mask detection process is divided into two distinct stages: first estimation (circular mask detection) and second estimation (elliptical mask detection). This segmentation allows each stage to focus on specific mask shapes, improving overall detection accuracy while keeping individual stages relatively simple.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The first estimation stage performs preliminary detection of the mask boundary assuming a circular shape. This preliminary action provides an initial approximation that is then refined in the second estimation stage, enabling accurate detection without requiring the second stage to handle all complexity alone.

Inventive Principle:

Principle #10Preliminary action

2Adaptability or versatility

If the mask position shifts frequently, then the adaptability of the detection system is improved, but the reliability of continuous accurate detection deteriorates

Engineering Contradiction:

Improvemask position adaptabilityVSAvoidcontinuous detection reliability

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The detection system dynamically adapts to mask position shifts by sequentially applying two different estimation models (circular then elliptical). This dynamic approach allows the system to handle various mask positions and shapes reliably, maintaining detection accuracy even when the mask shifts during observation.

Inventive Principle:

Principle #15Dynamics

3Ease of operation

If edge detection based on luminance values is used, then the ease of operation is improved, but the measurement precision of boundary points deteriorates

Engineering Contradiction:

Improveedge detection simplicityVSAvoidboundary point detection accuracy

Core Design Contradiction:

Ease of operationVSMeasurement precision

Solution Approach 1:

The system changes the estimation parameters between two stages: first assuming a circular shape with radius and center parameters, then assuming an elliptical shape with additional shape parameters. This parameter change approach refines boundary point precision while maintaining operational simplicity through automated parameter optimization.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10058237B2Image processing device, image processing method, and program

Publication Date: 2018.08.28 SONY GROUP CORP

AI summary

There is provided an image processing device including an edge detection unit configured to detect a boundary point between a first region including a subject to be observed and a second region that does not include the subject, a first estimation unit configured to estimate a first shape as a shape of a boundary between the first region and the second region based on the boundary point, and a second estimation unit configured to estimate a second shape as a shape of a boundary between the first region and the second region based on the boundary point and the estimated first shape.