Recovering Functional Mechanical Assemblies from Raw Scans.

Area of Science:

• Computer Vision
• Robotics
• Mechanical Engineering

Background:

• Reconstructing complex mechanical assemblies from raw 3D scan data presents significant challenges.
• Existing methods often struggle with part identification, precise positioning, and functional assembly.
• Automated reconstruction is crucial for applications in reverse engineering and digital manufacturing.

Purpose of the Study:

• To develop an automated method for reconstructing functional mechanical assemblies from multiple raw 3D scans.
• To accurately extract, parameterize, and position individual mechanical parts within an assembly.
• To enable the creation of a workable digital model suitable for physical fabrication.

Main Methods:

• A motion-guided, patch-based hierarchical registration and labeling algorithm is employed for functional part extraction.
• Extracted parts are parameterized, and their interrelationships are encoded using a graph structure.
• Joint optimization is utilized to determine the optimal geometry, placement, and orientation of each component.

Main Results:

• The method successfully reconstructs functional mechanical assemblies from diverse raw scan data.
• Accurate identification and spatial arrangement of functional parts are achieved.
• The reconstructed models are validated through physical fabrication, demonstrating their workability.

Conclusions:

• The proposed method offers a robust solution for reconstructing functional mechanical assemblies from 3D scans.
• This approach facilitates automated reverse engineering and digital twin creation.
• The validated results highlight the potential for practical application in manufacturing and design.