AI Surface Inspection Using Projected Laser Indicia
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine vision inspection systems struggle with efficiently inspecting large or moderately sized surfaces, especially when assembly rates are high and only small areas require inspection, due to computational costs and complexity in processing large images.
Innovation Solution
The system uses strategically projected laser indicia onto the inspection surface, combined with a convoluted neural network (CNN) trained by stored images of the light indicia, to simplify the inspection process by focusing on specific areas of interest defined by the laser boundary, reducing computational complexity and increasing accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine vision inspection using cameras and sensors is implemented to inspect whether a component has been properly assembled, then inspection accuracy is improved, but computational cost and processing time increase significantly when dealing with large images of very large objects
Solution Approach 1:
The patent divides the large inspection surface into multiple smaller image segments that can be processed independently and in parallel. This segmentation reduces the computational burden on each processing unit while maintaining overall inspection accuracy, directly resolving the contradiction between inspection accuracy and processing time for large objects.
Solution Approach 2:
The patent introduces a spatial dimension by projecting laser indicia onto the work surface to create a reference framework. This additional dimensional information allows the system to efficiently locate and process only relevant areas of interest within large images, reducing overall processing time while maintaining inspection accuracy.
2Measurement precision
If CNN algorithms are used to detect anomalies in images of parts under inspection, then inspection accuracy approaches human logic and accuracy, but computation cost limits their ability to process larger images of very large objects on industrial scales
Solution Approach 1:
The patent segments large inspection images into smaller manageable patches that can be processed by CNN algorithms. This reduces the computational memory requirements and processing complexity while maintaining anomaly detection accuracy, enabling industrial-scale deployment of CNN-based inspection systems.
Solution Approach 2:
The patent applies CNN algorithms selectively to only those regions of interest identified by laser indicia projection, rather than processing entire large images. This partial action approach reduces computation cost significantly while maintaining high anomaly detection accuracy in critical areas.
3Reliability
If high percentages of production must be inspected to meet ever-increasing quality standards, then manufacturing quality is improved, but inspection efficiency decreases making human inspection obsolete
Solution Approach 1:
The patent replaces human inspection with an automated system combining laser projection and image processing. This substitution enables high-volume inspection to meet quality standards while maintaining or improving productivity, as the automated system can process images much faster than human inspectors.
Solution Approach 2:
The patent performs preliminary processing by projecting laser indicia to identify areas of interest before full inspection. This preliminary action enables the system to focus computational resources on critical regions, improving overall inspection efficiency while maintaining high quality standards across large percentages of production.
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
This approach enables efficient and accurate inspection of large or moderately sized surfaces by reducing the computational burden and focusing analysis on specific areas of interest, thereby improving inspection efficiency and accuracy in mass production settings.
Implementation Method 1
a light source for projecting light indicia onto the component assembled to the workpiece
Implementation Method 2
An imager includes an image sensor system for imaging the workpiece and signaling a current image of the workpiece to the controller
Data Source
AI summary
A system for identifying accurate assembly of a component to a workpiece is disclosed. The system includes a light source for projecting light indicia onto the component assembled to the workpiece. A controller includes an artificial intelligence (AI) element defining a machine learning model that establishes a convoluted neural network trained by stored images of light indicia projected onto the component assembled to the workpiece. An imager includes an image sensor system for imaging the workpiece and signaling a current image of the workpiece to the controller. The machine learning model directs inspection of the workpiece to the light indicia imaged by said imager. The AI element determines disposition of the component disposed upon the workpiece through the neural network identifying distortions of the light indicia in the current image.