Research and Development

Computer Vision for Experimental Mechanics

VisionLab is committed to developing computer vision and experimental methods for precise measurement, contributing to the development of advanced materials, components, structural assemblies, and computational modelling itself.

 

As a partner in the industrial development of video extensometers and computer vision, we explore the possibilities of DIC measuring systems application and develop new methods enabling their easy and efficient use in engineering practice and academia.

 

VisionLab is equipped with modern measuring devices and optometric prototypes, used by students for their research and utilised for commercial research activities, thus contributing to the improvement of the current state of knowledge.

Video Extensometry and Stereoscopic Digital Image Correlation

The laboratory uses modern digital image correlation systems developed by the X-Sight company, including advanced video extensometers, for measuring displacements and deformations, and stereoscopic measuring systems for optometric analysis of non-planar objects and component testing.

 

As part of our research, we develop applications of optical measurements for newly developed materials and components in cooperation with industrial partners. We strive to increase the accuracy, reliability and versatility of measuring devices. We research their limits and push them to the limit of physical possibilities. We work to improve and automate software tools for measuring devices utilising artificial intelligence.

Residual Stress Measurement

The analysis of residual stress effect on component behaviour is beneficial in numerous industrial and scientific fields. This knowledge is crucial in investigations associated with fatigue life, fracture behaviour, corrosion resistance, or dimensional stability.

 

VisionLab uses the MTS 3000 test rig (supplied by Hottinger Baldwin Messtechnik) for residual stress measurement using the semi-destructive hole drilling method. The device has systems for controlled drilling and automatic evaluation of residual stresses.

 

In the framework of the ongoing research, an automated optical measuring technique is being developed to replace strain gauge rosettes used for determining residual stress by the hole-drilling method. The appropriate setting of the measuring chain is solved, including minimisation of noise, which significantly affects the measurement results.

DIC Microscopy

VisionLab’s activities are aimed at developing a stereomicroscopic 3D video extensometer for measuring nanometric deformations of test specimens using a stereomicroscope with a planachromatic lens and using the stereoscopic 3D DIC technique. The research concentrates on calibration techniques, accuracy maximisation and noise reduction. The aim is to develop a practical and accurate measuring system.

Vibrography Using Artificial Intelligence-Aided Computer Vision

VisionLab, in cooperation with X-Sight, is engaged in the research and development of an optometric device using machine learning for non-contact measurement and analysis of deformations in dynamic periodic and quasi-periodic processes using artificial intelligence, which allows their reconstruction from nonuniform sampling.

 

The measuring chain integrates a camera system, high-precision synchronization electronics, and DIC measuring software Alpha, which, with the help of machine learning, automates the technically demanding actions of the operator, such as the setting of tools used for measurement and the subsequent analysis of the obtained data.