GPU-accelerated reconstruction
The reconstruction core is fully GPU accelerated, with multithreaded I/O and overlapped CPU/GPU processing.
GPU-accelerated standalone CT reconstruction software developed by the HEPS X-ray microscopy imaging beamline team, supporting TIFF data, micro/nano CT, jitter correction, phase recovery and multiple iterative reconstruction methods.

MOCUPY is a GPU-accelerated standalone CT reconstruction application developed by the HEPS X-ray microscopy imaging beamline team with support from the HEPS project. It is the standalone edition of HEPSCT, written in Python with CUDA-accelerated core algorithms for fast offline reconstruction of X-ray CT data.
The current public release is V2.0 (2025.07), with an updated TIFF I/O library and interface optimizations.





The reconstruction core is fully GPU accelerated, with multithreaded I/O and overlapped CPU/GPU processing.
Automatic multi-scale rotation-axis correction, subpixel nanoscale CT stage-jitter correction and wavelet-based ring artifact removal.
Single- and multi-distance inline phase recovery with TIE, CTF, tiehom_Rytov, tiehom_Paganin and HIOER algorithms.
FBP, Grid, EM, SIRT and CGLS reconstruction, with Grid reconstructing a 2k×2k projection dataset in seconds.
| Version | Platform | File size | Published | Downloads | |
|---|---|---|---|---|---|
| 2.0.0 CTRecon_V2.0_202507.rar | Windows · x64 | 1.7 GB | 03/11/2026 | 15 |
Download links are time-limited and support resume; verify the SHA-256 checksum.
Contact the maintainers above for usage, adaptation or customisation.