A Dataset of Electrical Components for Mesh Segmentation and Computational Geometry Research

Area of Science:

• Computer Vision
• Robotics
• Manufacturing Engineering

Background:

• High-quality data is essential for automated and digitally assisted assembly processes.
• Existing datasets may not adequately represent the complexity and variety of electrical and electronic components for segmentation tasks.

Purpose of the Study:

• To introduce a novel, comprehensive dataset of triangle meshes for electrical and electronic components.
• To facilitate research in automated manufacturing, specifically in spatial assembly and segmentation tasks.

Main Methods:

• Scraping Computer Aided Design (CAD) models from the internet.
• Creating a dataset of 234 triangle meshes with labeled vertices, categories, and subcategories.
• Performing statistical analysis including shape, size, distribution, and inter-rater reliability.

Main Results:

• A versatile dataset of 234 labeled triangle meshes representing electrical and electronic components.
• Detailed statistical analysis of the dataset's properties.
• Demonstration of multimodal task applicability and suggested usage approaches.

Conclusions:

• The presented dataset is a valuable resource for advancing automated assembly and manufacturing.
• The multimodal nature of the dataset supports diverse research applications in computer vision and robotics.
• Further research into spatial assembly and component recognition is encouraged by this dataset.