Comparison of low cost 3D structured light scanners for face modeling

Area of Science:

• Computer Vision
• 3D Scanning Technology
• Biomedical Engineering

Background:

• Accurate 3D human face modeling is crucial for various applications.
• High-resolution 3D scanners are often expensive and inaccessible.
• Evaluating cost-effective alternatives is essential for broader adoption.

Purpose of the Study:

• To compare the accuracy of low-cost structured light scanners against a high-end reference system for 3D human face modeling.
• To assess the potential of affordable structured light systems in generating detailed facial geometries.
• To determine optimal data processing strategies for low-cost scanner outputs.

Main Methods:

• Three structured light scanner systems (one reference, two low-cost) were used to capture 3D data of two human faces.
• Commercial software was employed for point-cloud filtering, mesh generation, and hole filling.
• Open-source software (CloudCompare) was utilized for data evaluation and comparison.
• Various filtering and smoothing levels were applied to reference data for comparative analysis.

Main Results:

• Low-cost structured light scanners demonstrated considerable potential in capturing spatial and morphological information of human faces.
• The study identified optimal data reduction levels for the reference system to facilitate comparison with lower-resolution scanners.
• Results indicate that affordable systems can achieve viable accuracy for 3D face modeling.

Conclusions:

• Affordable structured light scanners are a promising technology for 3D human face modeling.
• These systems offer a cost-effective solution without significant compromise on essential spatial and morphological data.
• Further research can explore advanced processing techniques to enhance the output of low-cost 3D scanning systems.