DICStudio

DICStudio is a Windows-only, Python 3.8–based 2D digital image correlation (DIC) toolbox with a modern Qt GUI. It provides a full port and extension of the Ncorr 2D DIC software into Python, using a compiled backend (ncorr.pyd) plus additional preprocessing, automation, and post-processing tools.

DICStudio is an independent research project and is not officially affiliated with the original Ncorr authors. It is a Python port and extension built on top of their open-source work.

---

Key features

Ncorr-based DIC core

• Uses the original Ncorr DIC algorithms through a compiled Python extension: ncorr.pyd.
• Supports running full-field displacement and strain analysis.
• Can read and write Ncorr .bin files:
  • save_DIC_input, save_DIC_output
  • save_strain_input, save_strain_output
  • load_DIC_output, load_strain_output
• Long-running jobs run in a background QThread (NcorrWorker) with logging and progress reporting.

---

Preprocessing & experiment setup

The Preprocessing.MainWindow provides a multi-tab GUI for:

• Frame extraction (FrameExtractorTab)

  • Extract images from video files or sequences.
  • Manage reference and deformed frames.

• ROI definition & mask generation (ROITab, ROIDrawer)

  • Draw and edit regions of interest.
  • Generate binary ROI masks compatible with Ncorr.

• Calibration

  • Unit / scale calibration.
  • Camera-related utilities and pattern quality checks.

• Pattern generation & virtual experiments

  • Tools such as PatternGeneratorTab, PatternMaster, VirtualLabGeneratorTab.
  • Generate speckle patterns and perform virtual experiments / comparisons.

• Quality assessment & uncertainty

  • Tabs like QualityAssessmentTab, UQComparisonTab for evaluating pattern, ROI and analysis quality.

• Background removal (optional)

  • Integration with rembg for automatic background removal.
  • On first use, DICStudio can download the required ONNX model (~176 MB) with user consent.

---

Manual DIC mode

ManualModeWindow groups the main workflows in tabs:

• Preprocessing – all the ROI, pattern and calibration tools.
• DIC / strain analysis – the main processing window (MainWindow in main_app.py) to run Ncorr-based DIC and strain.
• Point inspection & analysis – PointInspectorApp for detailed post-processing.

This mode is intended for users who want full manual control over each step.

---

Automation controller

The AutomationController provides high-level automation:

• Automatically detects ROIs (e.g. using largest contour from frames).
• Prepares masks and launches Ncorr analysis in the background.
• Integrates with FrameExtractorTab, NcorrWorker and the GUI.
• Designed for batch / repeated experiments where the workflow is similar.

---

Point inspector & post-processing

point_inspector.py implements an extensive Point Inspector GUI (PointInspectorApp) with multiple analysis tabs, including:

• PointInspectorTab – inspect displacement/strain time series at selected points.
• RelativeDisplacementTab – relative displacements between points or regions.
• SDICalculatorTab – scalar metrics derived from DIC/strain data.
• PoissonRatioTab – compute and visualize Poisson-like quantities.
• AdvancedPlottingTab – flexible plotting (filters, smoothing, transforms).
• CMODTab – crack mouth opening displacement tools.
• FractureEnergyTab – fracture energy and related metrics.
• BinToCsvTab – convert Ncorr .bin files into CSV.

Most operations use NumPy/SciPy and Matplotlib, and results can be exported as CSV or image files.

---

Theming & launcher

• ThemeManager reads dark_theme.qss and light_theme.qss, detects system theme and applies appropriate styling.
• LauncherWindow provides a simple start screen:
  • Launch Manual DIC Mode.
  • Launch Automation / Batch Mode.
  • Open plotting / inspection windows.
• Custom icons:
  • logo.png, icon_dark.png, icon_light.png.

---

Repository layout

DICStudio/
  src/
    analysis_worker.py
    automatic_config_dialog.py
    automation_controller.py
    custom_widgets.py
    dialogs.py
    launcher.py
    main_app.py
    manual_mode_window.py
    plot_window.py
    point_inspector.py
    Preprocessing.py
    theme_manager.py
    __init__.py
  dependencies/
    ncorr.pyd           # compiled Ncorr backend for Python 3.8, Windows
    D3Dcompiler_47.dll
    libblas.dll
    libEGL.dll
    libfftw3-3.dll
    libgcc_s_sjlj-1.dll
    libgfortran-3.dll
    libGLESv2.dll
    liblapack.dll
    libquadmath-0.dll
    opencv_core300.dll
    opencv_highgui300.dll
    opencv_imgcodecs300.dll
    opencv_imgproc300.dll
    opencv_videoio300.dll
    python38.dll
    ncorr.exp
    ncorr.lib
    ncorr_lib.lib
  run_app.py
  requirements.txt
  dark_theme.qss
  light_theme.qss
  logo.png
  icon_dark.png
  icon_light.png
  force_data.csv
  LICENSE

The dependencies/ folder bundles the Windows-specific runtime libraries needed for the compiled ncorr.pyd and related functionality. These binaries are provided as part of this research prototype; if you intend to redistrib­ute them in other contexts (e.g. commercial software), you should review the corresponding third-party licenses.

---

Installation

1. Environment

• OS: Windows 10/11, 64-bit
• Python: 3.8.x only
  (ncorr.pyd and the DLLs in dependencies/ are built for Python 3.8 on Windows.)

Create and activate a virtual environment (recommended):

python -m venv .venv
.venv\Scripts�ctivate

2. Install Python dependencies

From the repository root:

pip install -r requirements.txt

This will install NumPy, SciPy, scikit-image, PySide6, OpenCV, rembg, Matplotlib and other required packages.

---

Running DICStudio

From the repository root:

python run_app.py

What this script does:

• Adds src/ and dependencies/ to sys.path.
• Configures Qt high-DPI behaviour.
• Creates a QApplication, applies the current theme, and shows LauncherWindow.

From the launcher you can:

• Start Manual DIC Mode.
• Start Automation / Batch Mode.
• Access analysis / plotting tools.

---

Basic workflow

1. Preprocessing

   • Load images or video, extract frames.
   • Define ROI and masks.
   • Calibrate units and camera if needed.
   • Optionally remove background and assess pattern / ROI quality.

2. Run DIC (Ncorr backend)

   • Configure subset size, step, strain radius and other parameters.
   • Start the analysis and monitor the progress dialog.
   • Save or load Ncorr .bin outputs as needed.

3. Post-processing

   • Use the Point Inspector to:
     • Select points or regions of interest.
     • Compute time-series, relative displacements, CMOD, fracture metrics, etc.
   • Export CSV files and plots for further analysis or publication.

---

Citing DICStudio and Ncorr

If you use DICStudio in academic work, please cite:

DICStudio (this project)

A. Mirjafari, “DICStudio: A Python 3.8 Port and Extension of Ncorr for 2D Digital Image Correlation,” [Journal/Conference], [Year].
(Update this section once the paper is accepted and has a final reference.)

Original Ncorr work

The original Ncorr project and its algorithms are described in:

J. Blaber, B. Adair, and A. Antoniou,
“Ncorr: Open-Source 2D Digital Image Correlation Matlab Software,”
Experimental Mechanics, 55(6), 1105–1122, 2015.
https://doi.org/10.1007/s11340-015-0009-1

and the accompanying code is hosted at:
https://github.com/justinblaber/ncorr_2D_cpp

---

License

DICStudio is released under the BSD 3-Clause License (see LICENSE).

• The Python source code in src/ and the project infrastructure are © Ali Mirjafari.
• The compiled backend ncorr.pyd and corresponding ncorr_lib artifacts are derived from the original Ncorr sources, which are also licensed under BSD 3-Clause. For full licensing details of Ncorr, please refer to the original Ncorr repository and documentation.

Third-party libraries (Python itself, FFTW, LAPACK, OpenCV, etc.) and their DLLs are subject to their own licenses. They are included here solely as runtime dependencies for research and reproducibility purposes.