Illusify is a mechanical virtual lab that generates synthetic image data for digital image correlation and computed tomography — from finite element simulations and 3D geometry models. Useful for education, algorithm development, and machine learning research.
Each module generates synthetic image sequences with corresponding ground-truth displacement and geometry data.
Surface displacement and strain fields
Generates speckle-pattern image sequences from 2D finite element results. Covers analytical kinematics and FE simulation import.
Stereo camera pair synthesis
Simulates a stereo camera setup from 3D FE models. Outputs synchronized image pairs for stereo DIC analysis pipelines.
X-ray projections from 3D geometry
Generates X-ray computed tomography projection sequences from STL models. Beer-Lambert attenuation with configurable multi-material support.
Configure a test in the dashboard or use the MCP server to let an AI assistant set parameters and submit jobs on your behalf.
2D-DIC, stereo 3D-DIC, CT radiography — or upload your own FE results.
Define geometry, speckle pattern, strain range, noise level, and number of frames.
Check the reference image before running the full sequence.
The server renders all frames and packages results as a downloadable ZIP.
Feed images to your DIC software or ML pipeline. Ground-truth fields are included.
Students can carry out virtual mechanical tests — including standardized dogbone tensile tests — without physical lab equipment. The full image sequence can be processed with standard DIC software, giving an experience close to a real experiment. Educators can configure test scenarios and verify results against known ground truth.
The traditional workflow runs a physical test, extracts DIC data, fits a material model — and hopes the pipeline is right. Illusify reverses this: define your material model, generate a synthetic test with known ground truth, and validate the entire characterisation pipeline — DIC software, FEMU routines, and VFM solvers — against a result you already know.
Labeled synthetic data with full ground-truth displacement, strain, and geometry fields is hard to obtain from physical experiments. Illusify lets you generate as much as you need — with controlled parameters, reproducible conditions, and no physical throughput constraint.
We're currently working with a small group of invited users while we refine the platform. If you'd like to request access or learn more, reach out and we'll get back to you.
info@illusify.no