Railway Facility Inspection Using Point Cloud Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated inspection systems for railroad facilities struggle to detect foreign objects and damages outside the operating vehicle area due to alignment deviations caused by moving objects, leading to false detection of normal parts as abnormal.
Innovation Solution
A facility inspection system and method that includes a photographing device, storage device, alignment area separation unit, alignment unit, and difference extraction unit, which separates and aligns three-dimensional point clouds to exclude moving objects from the alignment area, allowing for accurate detection of abnormalities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated inspection using three-dimensional point cloud comparison is implemented, then inspection frequency and productivity are improved, but measurement precision deteriorates due to alignment deviations caused by moving objects
Solution Approach 1:
The patent segments the three-dimensional point cloud into multiple regions (alignment region and non-alignment region) based on their functional roles. The alignment region contains only stationary objects used for registration, while the non-alignment region contains the inspection target. This segmentation prevents moving objects from interfering with the alignment process, thereby maintaining measurement precision while enabling automated high-frequency inspection.
Solution Approach 2:
The patent extracts and isolates the alignment region from the complete three-dimensional point cloud data. By separating the alignment region (containing only stationary objects) from the non-alignment region (containing inspection targets and moving objects), the system eliminates the harmful effect of moving objects on alignment accuracy, allowing precise measurement despite automated inspection frequency.
2Ease of manufacture
If automated inspection systems are deployed, then labor cost is reduced, but detection accuracy worsens due to false detection of normal parts as abnormal
Solution Approach 1:
The patent divides the point cloud into alignment region and non-alignment region, ensuring that only stationary objects in the alignment region are used for registration. This prevents moving objects from causing alignment deviations that would lead to false detections, thereby maintaining high detection accuracy while enabling automated inspection that reduces labor costs.
Solution Approach 2:
The patent introduces the alignment region as an intermediary component that mediates between the automated inspection system and the inspection target. This intermediary layer filters out moving objects and provides a stable reference basis for alignment, preventing false detections while maintaining the benefits of automated inspection.
3Device complexity
If alignment is performed using barycentric position comparison, then the process is simplified, but reliability deteriorates when moving objects are present in the point cloud
Solution Approach 1:
The patent segments the point cloud to create a dedicated alignment region containing only stationary objects. This segmentation maintains the simplicity of the barycentric alignment process while improving reliability by ensuring that moving objects do not contaminate the alignment calculation, thus achieving both simplicity and reliability.
Solution Approach 2:
The patent applies different quality requirements to different regions: the alignment region is designed to contain only stationary objects with high stability for reliable alignment, while the non-alignment region can contain various objects including moving ones. This local quality differentiation ensures alignment reliability without complicating the overall process.
Data Source
AI summary
A facility inspection system prevents a normal part from being detected as an abnormal part caused by a deviation in an alignment due to a presence/absence of a moving object in detecting the abnormal part in a surrounding environment of a vehicle moving on a track. The system includes a photographing device, storage device, separation unit, an alignment unit, and a extraction unit. The photographing device photographs the surrounding environment of the moving vehicle. The storage device stores a reference alignment point cloud and a reference difference-extraction point cloud for each position on the track. The separation unit separates the alignment point cloud from a three-dimensional point cloud. The alignment unit aligns the reference alignment point cloud and the alignment point cloud and outputs alignment information. The extraction unit extracts a difference between the three-dimensional point cloud deformed based on the alignment information and the reference difference-extraction point cloud.