3D Point Cloud Profile Matching for Industrial As-Built Plans
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for obtaining 'as-built' plans of industrial installations, such as those using laser scanning, produce three-dimensional clouds of points that are difficult to convert into usable computer data for rigorous planning and design, especially for recognizing regular shapes like elongated profiles.
Innovation Solution
A method and apparatus for processing image data by storing three-dimensional clouds of points, selecting a two-dimensional profile, and matching it with the cloud points within regions to obtain sets of data representing matched surface portions, allowing for precise and fast recognition of regular shapes, including the derivation of three-dimensional curves from matched surface portions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If laser scanning is used to obtain three-dimensional clouds of points, then the general contour of the installation can be captured, but the cloud of points is hardly usable for rigorous planning and designing
Solution Approach 1:
The patent segments the cloud of points into multiple regions and divides the profile matching process into incremental steps. Each step matches the profile to a subset of points within a defined region, building up the complete matched surface portion by combining results from multiple segments. This segmentation enables rigorous planning while maintaining operational feasibility through automated processing.
Solution Approach 2:
The patent introduces an intermediary processing stage that converts the raw cloud of points into processed data representing matched surface portions. This intermediary transformation layer translates the general contour information into structured data that can be directly used for planning and design, bridging the gap between measurement capability and operational usability.
2Manufacturing precision
If the profile matching process covers the entire cloud of points at once, then complete coverage is achieved, but the processing time becomes excessively long
Solution Approach 1:
The patent divides the cloud of points into multiple regions and processes the profile matching incrementally across these regions. Each region is handled in separate processing steps, allowing the system to achieve complete coverage through combination of partial results while maintaining fast processing speeds. This segmentation strategy enables both high precision and productivity.
Solution Approach 2:
The patent performs preliminary actions by defining regions and establishing matching parameters before executing the full profile matching process. The system pre-divides the cloud into manageable regions and prepares the matching algorithm to process each region efficiently, enabling fast completion while maintaining comprehensive coverage and precision.
3Ease of operation
If the cloud of points is processed to recognize regular shapes, then design usability is improved, but the complexity of data processing increases
Solution Approach 1:
The patent segments the complex profile matching process into simpler, manageable steps that operate on regional subsets of points. Each step involves matching the profile to points within a defined region, which simplifies the overall complexity while enabling comprehensive shape recognition. This segmentation makes the process more operationally feasible despite the inherent complexity of recognizing regular shapes in point clouds.
Solution Approach 2:
The patent introduces intermediary processing stages that simplify the transformation from raw point cloud data to recognized shape data. These intermediary steps break down the complex recognition task into manageable operations, reducing the overall processing complexity while maintaining the ability to identify regular shapes for design purposes.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables fast and precise conversion of three-dimensional clouds of points into computer data representing drawings or plane elements, facilitating the recognition of regular shapes and improving the usability of 'as-built' plans for design and planning purposes.
Implementation Method 1
At each laser position, the distance is derived from the time or the phase difference in the laser to-and-fro path to a reflecting point
Data Source
AI summary
Methods and devices for processing image data stored as three-dimensional point clouds related to a scene by selecting a profile definable in a two-dimensional surface, and matching the profile to one or more subsets of the cloud of points to identify surfaces in the scene having the same profile so that data sets representing the successive matched surface portions can be generated to provide a surface model of all or part of the scene.