Welding Image Defect Detection Using Weld Pool Validation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing defect detection technologies in welding require complex database updates due to changes in welding objects, parameters, or imaging conditions, and struggle with reliability when defects are obscured by blown-out highlights or shadows.
Innovation Solution
A processing device uses a first model to directly detect weld pools and defects from welding images, determining the appropriateness of defect detection results based on weld pool detection, thereby eliminating the need for a feature-defect relationship database and enhancing user convenience while improving reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a feature-defect relationship database is used for defect detection, then defect detection can be performed based on extracted features, but the system requires complex database updates when welding objects, parameters, or imaging conditions change
Solution Approach 1:
The patent extracts only the essential detection functionality from the complex feature-defect relationship database system. By using a detection model that directly identifies defects from images without requiring comprehensive feature databases, the system removes the burden of database maintenance and updates while preserving defect detection capability
Solution Approach 2:
The detection model performs self-adaptation to different welding conditions without requiring external database updates. The model learns from training data and can generalize to new welding objects, parameters, and imaging conditions autonomously, eliminating the need for manual database curation and updates
2Adaptability or versatility
If defect detection is performed in challenging imaging conditions with blown-out highlights or shadows, then comprehensive quality assessment is achieved, but detection reliability decreases due to obscured defects
Solution Approach 1:
The system performs preliminary processing of the welding image including normalization and enhancement before defect detection. By pre-processing the image to reduce the impact of blown-out highlights and shadows, the system prepares optimal input conditions for the detection model, improving reliability in challenging imaging conditions
Solution Approach 2:
The patent changes the parameters of the imaging data through normalization and enhancement techniques. By adjusting image parameters such as brightness, contrast, and color distribution, the system makes defects more visible under varying imaging conditions while maintaining detection accuracy
Data Source
AI summary
According to one embodiment, the processing device acquires a first detection result and a second detection result by inputting a first image to a first model. The first model detects a welding element and a defect according to an input of a welding image. The first image is imaged when welding. The first detection result relates to the welding element. The second detection result relates to the defect. The processing device determines an appropriateness of the second detection result by using the first detection result.