3D Vision Inspection for Industrial Defect Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current industrial computer vision systems rely on 2D imaging for inspection, which is inadequate for features characterized by physical size, area, or volume, as it loses critical 3D shape information and can be biased towards visible features, whereas 3D imaging is not widely adopted due to the unreliability of consumer-grade 3D cameras.

Innovation Solution

A method for direct 3D vision inspection using 3D vision systems, involving a training phase to create a 3D alignment model and variance model, and a runtime phase for defect detection by aligning and comparing 3D test images with a reference image, utilizing 3D cameras and image filtering techniques to handle noise and occlusion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If 2D imaging is used for inspection, then the system is simple and cost-effective, but critical 3D shape information is lost and inspection results are biased towards visible features

Engineering Contradiction:
Improve3D shape informationVSAvoidimaging system complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent transitions from 2D imaging to 3D imaging by capturing depth information alongside intensity data. The 3D vision system acquires both 2D intensity images and corresponding depth maps, enabling inspection of features characterized by physical size, area, or volume without losing critical 3D shape information. This dimensional enhancement resolves the information loss problem while maintaining system feasibility.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Reliability

If consumer-grade 3D cameras are used, then the system becomes more capable of capturing 3D information, but reliability and stability are insufficient for industrial applications

Engineering Contradiction:
Improvecamera reliabilityVSAvoid3D shape information
Core Design Contradiction:
ReliabilityVSLoss of information

Solution Approach 1:

The patent changes the key parameter from consumer-grade 3D cameras to industrial-grade 3D cameras, which provide the necessary reliability and stability for industrial applications. Industrial cameras offer improved noise characteristics, better depth accuracy, and enhanced robustness while maintaining the capability to capture comprehensive 3D information including shape, size, and volume features.

Inventive Principle:
Principle #35Parameter changes

3Ease of operation

If 3D images are converted to 2D images for inspection, then the inspection process is simplified, but information about 3D shape or volume is lost and results are biased

Engineering Contradiction:
Improveinspection process simplicityVSAvoid3D shape and volume information
Core Design Contradiction:
Ease of operationVSLoss of information

Solution Approach 1:

The patent maintains 3D image data throughout the inspection process rather than converting to 2D. The system performs registration, comparison, and defect detection directly on 3D point clouds and depth maps, preserving all 3D shape and volume information. This approach enables accurate measurement of features characterized by physical dimensions while avoiding the bias inherent in 2D projections.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

4Measurement precision

If direct 3D vision inspection is implemented, then accurate detection of 3D defects is enabled, but system complexity and processing requirements increase

Engineering Contradiction:
Improvedefect detection accuracyVSAvoidvision system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the inspection process into distinct phases: data acquisition (intensity and depth), registration (alignment to CAD model), comparison (defect detection), and classification. This segmentation manages complexity by organizing processing steps while enabling precise 3D defect detection through specialized algorithms for each phase, including point-to-surface distance calculations and statistical analysis.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP3496035B1Using 3D vision for automated industrial inspection
Publication Date: 2020.12.09 COGNEX CORP
  • EP3496035B1 patent drawingFigure 1
  • EP3496035B1 patent drawingFigure 2
  • EP3496035B1 patent drawingFigure 3

AI summary

A system and method for three dimensional (3D) vision inspection using a 3D vision system. The system and method comprising acquiring at least one 3D image of a 3D object using the 3D vision system, using the 3D vision system; extracting a 3D visible runtime mask of the 3D image; using the 3D vision system, comparing the 3D runtime visible mask to a 3D reference visible mask; and, using the 3D vision system, determining if a difference of pixels exists between the 3D runtime visible mask and the 3D reference visible mask.