Methods, systems, media and equipment for inspecting profile cross-sectional dimensions
By using point cloud data processing and generative adversarial networks, burrs on profile cross-sections are identified and removed, solving the problems of low detection efficiency and insufficient accuracy in existing technologies, and achieving high-precision detection of profile cross-section dimensions.
Patent Information
- Application Number
- CN202610008236.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-05
- Publication Date
- 2026-05-26
AI Technical Summary
In the current technology for inspecting the quality of profile cross sections, manual inspection is inefficient, visual recognition accuracy is greatly affected by image quality, burrs affect the accuracy of dimensional measurement, and two-dimensional imaging cannot accurately locate the position, resulting in large deviations in measurement results.
By employing point cloud data processing and generative adversarial networks, and through point cloud registration, segmentation, and burr synthesis detection models, the influence of burrs is identified and eliminated. The burr synthesis detection model is trained using generators and discriminators to generate realistic burr masks. Combined with a burr size recognition model, the detection accuracy is improved.
It improves the accuracy of profile cross-sectional dimension inspection, eliminates the influence of burrs, enhances the precision and robustness of inspection results, and solves the problem of burrs interfering with measurements.
Smart Images
Figure CN122089648A_ABST