Surface Abnormality Detection Using Local Roughness Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing abnormality detection systems inaccurately detect surface abnormalities in structures with varying curvatures due to large errors between shape data and reference shapes, particularly in tunnels with non-constant curvature.
Innovation Solution
An abnormality detection system that includes shape data acquisition, reference shape calculation, abnormality candidate extraction, degree of roughness calculation, and abnormality determination, utilizing a two-stage process to reduce erroneous detections by first identifying candidate areas based on shape data differences and then determining abnormalities based on surface roughness thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If shape data is compared with a reference shape to detect abnormalities, then abnormality detection capability is improved, but measurement precision deteriorates due to large errors in structures with varying curvatures
Solution Approach 1:
The patent segments the surface into multiple regions and divides the abnormality detection into two stages: first extracting abnormality candidate areas based on shape data differences, then determining actual abnormalities by calculating degree of roughness in those segmented regions. This segmentation approach allows the system to handle structures with varying curvatures by processing them in manageable portions rather than as a whole.
Solution Approach 2:
The patent applies local quality analysis by calculating the degree of roughness specifically in the abnormality candidate areas extracted in the first stage, rather than uniformly analyzing the entire surface. This localized approach enables the system to distinguish between normal curvature variations and actual abnormalities by examining the local surface quality characteristics of suspected regions.
2Device complexity
If a single-stage abnormality detection method is used, then device complexity is reduced, but measurement precision deteriorates leading to erroneous detections
Solution Approach 1:
The patent implements preliminary action by first extracting abnormality candidate areas based on shape data differences before performing the final abnormality determination. This two-stage process where the first stage prepares and narrows down candidate regions, and the second stage performs precise roughness analysis, improves measurement precision without requiring overly complex device architecture.
Data Source
AI summary
An abnormality detection system includes: shape data acquisition means for acquiring shape data of a surface of a structure; reference shape calculation means for calculating a reference shape for the shape data acquired by the shape data acquisition means; abnormality candidate extraction means for extracting an abnormality candidate area of the surface based on a difference between the reference shape calculated by the reference shape calculation means and the shape data corresponding to the reference shape; degree of roughness calculation means for calculating a degree of roughness of the abnormality candidate area extracted by the abnormality candidate extraction means and an area near the extracted abnormality candidate area; and abnormality determination means for determining that there is an abnormality in an area where the degree of roughness calculated by the roughness calculation means exceeds a threshold among the abnormality candidate areas extracted by the abnormality candidate extraction means.