Digital Inspection Technology for Sheet Metal Parts Using 3D Point Clouds.

Area of Science:

• Manufacturing Engineering
• Computer Vision
• Metrology

Background:

• Traditional sheet metal measurement methods suffer from low efficiency.
• Accurate and efficient digital inspection of sheet metal parts is crucial for quality control.
• Existing 3D point cloud processing techniques require optimization for feature preservation and registration accuracy.

Purpose of the Study:

• To develop a novel digital inspection method for sheet metal parts based on 3D point cloud data.
• To enhance the efficiency and accuracy of sheet metal measurement.
• To provide a reliable digital inspection solution for manufacturing quality control.

Main Methods:

• Acquisition of 3D point cloud data using a 3D laser scanner.
• Point cloud denoising (pass-through filtering) and downsampling (improved voxel grid).
• Feature extraction (ISS), description (FPFH), registration (SAC-IA, improved ICP), and surface reconstruction (greedy projection triangulation with MLS).

Main Results:

• The proposed method successfully processes 3D point cloud data for sheet metal inspection.
• Registration accuracy and efficiency are enhanced through improved ICP algorithm.
• Measurement errors for three workpieces ranged from 0.1416 mm to 0.2684 mm, meeting the ±0.3 mm requirement.

Conclusions:

• The digital inspection method based on 3D point clouds offers a reliable and accurate solution for sheet metal parts.
• The proposed techniques for point cloud processing and registration significantly improve measurement efficiency and accuracy.
• This approach meets the stringent accuracy requirements for industrial sheet metal inspection.