3D Point Cloud Flattening for Curved Surface Defect Measurement

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing machine vision systems struggle to reliably detect and evaluate defects and features on curved surfaces, such as cylinders and spheres, due to inefficiencies in mesh/surface reconstruction and sensitivity to lighting conditions, particularly when dealing with small or minute features.

Innovation Solution

A system and method that directly processes unorganized 3D point clouds by flattening them along cylindrical or spherical surfaces, allowing for blob detection and feature extraction without requiring mesh or voxel grid representations, and converting back to the original 3D space for accurate measurements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If mesh/surface reconstruction is used to detect blobs on curved surfaces, then detection capability is improved, but processing efficiency deteriorates

Engineering Contradiction:
Improvedetection capabilityVSAvoidprocessing efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent extracts only the essential geometric information needed for blob detection by directly processing point cloud data to identify surface deviations, rather than performing complete mesh reconstruction. This extraction approach maintains detection reliability while significantly reducing processing complexity and time.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the curved surface into local regions and processes each region independently to detect blobs. By dividing the complex reconstruction task into smaller localized operations, the system achieves reliable detection without the computational burden of full surface reconstruction.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If mesh/surface reconstruction is performed, then surface representation accuracy is improved, but processing time increases

Engineering Contradiction:
Improvesurface representation accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary filtering and organization of point cloud data before blob detection, pre-processing the data to enhance surface representation accuracy. This preliminary action reduces the need for time-consuming post-processing and full mesh reconstruction while maintaining measurement precision.

Inventive Principle:
Principle #10Preliminary action

3Device complexity

If traditional blob analysis tools referencing planar surfaces are used, then processing simplicity is maintained, but detection reliability on curved surfaces deteriorates

Engineering Contradiction:
Improveprocessing simplicityVSAvoiddetection reliability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent applies local quality analysis by adapting the reference surface to match the local curvature at each detection point. Instead of using a uniform planar reference, the system dynamically adjusts the reference geometry to the local surface characteristics, maintaining processing simplicity while significantly improving detection reliability on curved surfaces.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP4229588B1System and method for extracting and measuring shapes of objects having curved surfaces with a vision system
Publication Date: 2026.03.25 COGNEX CORP
  • EP4229588B1 patent drawingFigure 1
  • EP4229588B1 patent drawingFigure 2
  • EP4229588B1 patent drawingFigure 3~4

AI summary

This invention provides a system and method that efficiently detects objects imaged using a 3D camera arrangement by referencing a cylindrical or spherical surface represented by a point cloud, and measures variant features of an extracted object including volume, height, and center of mass, bounding box, and other relevant metrics. The system and method, advantageously, operates directly on unorganized and unordered points, requiring neither a mesh/surface reconstruction nor voxel grid representation of object surfaces in a point cloud. Based upon a cylinder/sphere reference model, an acquired 3D point cloud is flattened. Object (blob) detection is carried out in the flattened 3D space, and objects are converted back to the 3D space to compute the features, which can include regions that differ from the regular shape of the cylinder/sphere. Downstream utilization devices and/or processes, such as part reject mechanism and/or robot manipulators can act on the identified feature data.