multimodal-registration

A Python package for registering and combining 3-D volumes from different synchrotron imaging modalities — primarily Diffraction Contrast Tomography (DCT) and Phase Contrast Tomography (PCT) — and for propagating Digital Volume Correlation (DVC) displacement fields into the crystallographic data.

---

What it does

Step	Description
Load	Read DCT (HDF5), PCT (TIFF / HDF5), and DVC (VTK / MAT) volumes
Register	Upscale and pad DCT to the PCT voxel grid, then align using phase cross-correlation
Inspect	Interactive overlay for manual fine-tuning (Jupyter or script)
Deform	Warp DCT volumes with DVC displacement fields
Rotate	Update IPF colour maps to reflect DVC lattice rotations

---

Quick start

from multimodal_registration import DCT, ReferencePCT, DVC
from multimodal_registration import register, apply_dvc

# 1 — load
dct = DCT("scan.h5")
pct = ReferencePCT("reference.tiff")
dvc = DVC("results.vts", displacement_key="displacement",
          strain_keys=["strain"])

# 2 — bring DCT into PCT frame
dct.flip(axis=0)                   # match orientation
dct.upscale(pct_vox_size)          # match voxel size
dct.pad(pct.shape)                 # match spatial extent

# 3 — automatic registration
shift = register(dct, pct)

# 4 — apply DVC deformation + update IPF
apply_dvc(dct, dvc, crystal_symmetry="cubic")

---

Package layout

multimodal_registration/
├── backends.py       GPU / CPU detection (CuPy → NumPy fallback)
├── dct.py            DCT class  — load, upscale, pad, shift, flip
├── pct.py            ReferencePCT class — load, Otsu mask, binary closing
├── dvc.py            DVC class — VTK and MATLAB/HDF5 loaders
├── registration.py   Phase cross-correlation registration
├── interactive.py    ManualRegistration — Jupyter widgets / keyboard UI
└── deformation.py    Warp + IPF update from DVC displacement fields