Point Cloud Analysis for Linear Structure Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Laser measurement of elongated structures like branch lines or cables faces challenges due to varying point cloud density, leading to detection failures and false model estimations, especially when the structure's width is narrower than the laser scan interval, causing data loss and accuracy issues.
Innovation Solution
A point cloud analysis device and method that estimates the presence or absence of linear structures by utilizing the length and relationship between divided regions of the structure, projecting point clouds along the central axis to calculate an evaluation score robustly, independent of point cloud density, and incorporating local geometric information to exclude outliers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If laser measurement is performed on elongated structures with narrow width, then the structure can be captured in the measurement range, but the point cloud density becomes extremely sparse leading to detection failure
Solution Approach 1:
The patent changes the evaluation parameter from point cloud density to the area of the model surface. By evaluating the area of the fitted model surface rather than the density of points, the system can accurately detect linear structures even when point cloud density is extremely low, thus resolving the contradiction between detection accuracy and point cloud quantity
Solution Approach 2:
The patent introduces an area-based evaluation function as an intermediary between the raw point cloud data and the final detection result. This intermediary metric (model surface area) provides a stable basis for evaluation that is independent of point cloud density variations, enabling reliable detection across different measurement conditions
2Area of stationary object
If the measurement area or number of point clouds changes due to relative position and posture, then the measurement coverage varies, but the point cloud density changes causing difficulty in setting evaluation scores
Solution Approach 1:
The patent changes the evaluation parameter from point cloud density to model surface area. This parameter transformation makes the evaluation score independent of measurement area and point cloud density variations caused by changes in relative position and posture, ensuring consistent evaluation across different measurement conditions
Solution Approach 2:
The patent creates an equipotential evaluation system where the model surface area metric provides a uniform basis for evaluation regardless of the measurement conditions. By using area rather than density, the system achieves evaluation consistency across varying measurement areas and point cloud quantities
3Measurement precision
If RANSAC is used for robust estimation, then accurate shape can be estimated with many inlier points, but false detection models occur when ratio of outlier points is high
Solution Approach 1:
The patent changes the evaluation parameter from point cloud density to model surface area. This parameter change provides a more reliable basis for distinguishing true models from false detections, as the area metric remains stable even when the ratio of outlier points is high, thus reducing false detection rates while maintaining estimation accuracy
Data Source
AI summary
Provided is a point cloud analysis device that curbs a decrease in model estimation accuracy due to a laser measurement point cloud. A clustering unit (30) clusters a point cloud representing a three-dimensional point on an object obtained by a measurement unit mounted on a moving body and performing measurement while scanning a measurement position, within a scan line, to obtain a point cloud cluster. A central axis direction estimation unit (32) estimates a central axis direction based on the point cloud cluster. A direction-dependent local effective length estimation unit (34) estimates a local effective length based on an estimated central axis direction and an interval of scan lines, the local effective length being a length when a length of projection of the point cloud cluster in a central axis direction for each of the point cloud clusters is interpolated by an amount of a loss part of the point cloud.