Feasibility Study of a Motion Capture Based, Moveable Three-Dimensional Digital Image Correlation System
Year: 2025
Type: conference-paper
Abstract: This paper discusses the development and testing of a cost-effective and portable three-dimensional Digital Image Correlation (3D DIC) system optimized for larger specimen deformation monitoring and analysis. Traditional 3D DIC systems face limitations in scalability due to calibration constraints and sensitivity and can be expensive. A novel dual-camera mount was designed, manufactured and tested to address these issues, allowing for simplified calibration and enhanced data accuracy by allowing for increased control over the angle between the cameras and their position relative to the specimen using motion capture technology. Initial results using a MATLAB-based DIC analysis framework demonstrated improved accuracy in image mapping, despite challenges with initial designs of the camera mount which followed more traditional designs where camera angles are set for each individual camera, leading to increased chances of poor calibration. Results are compared against the pre-calibrated VIC 3-D software developed by Correlated Solutions for comparison and validation, aiming to achieve accuracy within acceptable error margins for industrial application. This paper will also show the improvements made to the system throughout testing, including upgrades to the camera mount to improve stability and optimizations to the software to improve data analysis. The system outlined within this paper has the potential to allow users to move and scan around a larger specimen due to the camera mount that was designed and experimentation that determined optimum distances and angles for data collection. This potential to move the 3D DIC system without needing to calibrate the system after each movement sets it apart from the VIC 3-D software that while more accurate after an initial calibration becomes virtually unusable after even the slightest adjustment is made to camera position. The future potential of an easily moveable 3D DIC system will also be discussed within this paper.
DOI: 10.2514/6.2025-98947
Authors: Andrew Walters, Nathan Snyder, Elise Roberson, Grant Custer, Jichul Kim
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4413191005