Automated Visual Inspection Using Golden Sample Defect Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automated visual inspection techniques in manufacturing are prone to false positives and false negatives, and struggle with detecting anomalies that do not rise to the level of defects, particularly in environments where articles can become dirty during processing.
Innovation Solution
A method and system that utilizes an object detection model and a golden sample image to verify defect detections, with a user interface for confirming anomalies, and a generative machine learning model to generate a golden sample image from an inspection image, allowing for retraining of the object detection model and improved defect classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated visual inspection techniques are used to increase inspection speed and accuracy, then productivity and measurement precision are improved, but false positives and false negatives increase, reducing reliability
Solution Approach 1:
The patent introduces an intermediary verification step where detected defects are compared against a golden sample image to confirm whether they represent actual defects or artifacts. This intermediary comparison mechanism resolves the contradiction by filtering out false positives while maintaining high inspection speed through automated processing.
Solution Approach 2:
The system implements feedback through the comparison process where detection results are validated against reference data. The feedback loop confirms or refutes initial detections, improving reliability without sacrificing the high-speed automated inspection capability.
2Speed
If traditional automated visual inspection is implemented, then inspection speed increases, but the system cannot distinguish between defects and anomalies such as dirt or processing marks
Solution Approach 1:
The golden sample image serves as an intermediary reference that enables precise differentiation between defects and anomalies. By comparing detected features against this reference, the system maintains high inspection speed while achieving accurate defect classification.
Solution Approach 2:
The system changes the parameter of comparison by introducing a reference-based verification step that transforms the detection process from simple anomaly identification to precise defect classification, enabling distinction between actual defects and processing artifacts.
3Measurement precision
If manual visual inspection is used to improve detection accuracy, then measurement precision improves, but productivity and inspection speed decrease
Solution Approach 1:
The patent replaces the mechanical human inspection process with an automated computer vision system that uses image comparison algorithms. This substitution maintains high detection accuracy through systematic comparison while dramatically increasing inspection speed and productivity.
Solution Approach 2:
The system creates a digital copy (golden sample image) of the defect-free or reference state, which is then used for automated comparison. This copying approach enables rapid automated verification that matches or exceeds human accuracy while achieving much higher inspection throughput.
4Measurement precision
If hardware requirements are increased to improve automated inspection capability, then measurement precision improves, but device complexity and cost increase
Solution Approach 1:
The patent replaces complex hardware solutions with a software-based image comparison approach. By using computational methods to compare inspection images against golden samples, the system achieves high detection accuracy without requiring additional complex hardware components.
Solution Approach 2:
The system uses digital copying and processing of images rather than physical hardware modifications. The golden sample image serves as a digital reference that enables precise detection through software comparison, avoiding the need for additional sensors, cameras, or optical components.
Data Source
AI summary
A system and method for automated visual inspection is provided herein. The method includes providing an inspection image of the article to an object detection model trained to detect at least one defect type in an input image and generating object location data identifying a location of a detected object in the inspection image; comparing the inspection image to a golden sample image to identify an artifact in the inspection image corresponding to a difference between the inspection image and the golden sample image, wherein the artifact is defined by artifact location data describing a location of the artifact in the inspection image; and determining whether the artifact location data matches the object location data according to predetermined match criteria.