Skip to content
Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Space-time Tomography for Continuously Deforming Objects

This is the open source repository for paper in SIGGRAPH 2018:

Space-time Tomography for Continuously Deforming Objects

Guangming Zang, Ramzi Idoughi, Ran Tao, Gilles Lubineau, Peter Wonka and Wolfgang Heidrich

King Abdullah University of Science and Technology (KAUST)


X-ray computed tomography is a valuable tool for analyzing objects with interesting internal structure or complex geometries that are not accessible with optical means. Unfortunately, tomographic reconstruction of complex shapes requires a multitude (often hundreds or thousands) of projections from different viewpoints, which can only be achieved with mechanical motion. This significantly limits the ability to use x-ray tomography for either objects that undergo uncontrolled shape change at the time scale of a scan, or else for analyzing dynamic phenomena, where the motion itself is under investigation


The code is tested in Visual Studio 2015 and 2018 on 64 bits Windows 7 and Windows 10. All the input raw data and experimental results can be found here

To run the ST tomography framwork, you need to install argtable, openmp, eigen, and cimg libraries

For command and parameter usage, you can find more detail in the paper and supplement, or simply use ST-Tomography --help to find more detail:

  -s, --Sigma=<double>                         Sigma in volume density reconstruction (default as 0.2)
  -t, --Tau=<double>                           Lambda in volume density reconstruction (default as 0.2)
  -l, --Lambda=<double>                        TV prior weight in volume density reconstruction
  -k, --TemporalPrior=<double>                 Temporal prior weight in volume density reconstruction
  -e, --Bright constancy prior=<double>        Bright constancy weight in volume density reconstruction
  --huber=<double>                             Trade-off weight for huber norm epsilon
  -a, --Alpha=<double>                         Stepsize for SART algorithm  (default as 0.3)
  -o <output>                                  Output 3d image (default as "hello.tif")
  -i, --projsvolume=<file>                     Input Dataset file(.mha.tif.tiff 2d bmp and other raw file are supported)
  -f, --XYZ=<int>                              Takes an integer value (defaults to 9)
  -u, --imgUV=<int>                            w and h of projection image default:512
  -c, --voxelspacing=<double>                  Voxel spacing (default as 1)
  -d, --detector spacing=<double>              Detector spacing (default as 1)
  -g, --sid=<double>                           Source to Iso-Object Distance  (default as 600mm)
  -j, --sdd=<double>                           Source to Detector Distance (default as 1000mm)
  --oxyz=<double>                              Offset x y z of the center of volume (default as 0)
  --sdg=<double>                               Start degree for each proj sequence
  -v, --nframes=<int>                          Number of frames
  -r, --nframes=<int>                          Number of rounds for each frames
  -b, --AlgorithmIter=<int>                    Number of algorithm iterations, default as 20
  -p, --SartIter=<int>                         Number of SART nested iterations, default as 1
  --prior={STV,SAD,ATV}                        Specify the prior you are applying from {STV,SAD,TV} , default as TV
  --bp={Voxelbased,Raybased}                   Specify the backprojection mode {Voxelbased,Raybased}, default as voxelbased        
  -h, --help                                   Print this help and exit
  --version                                    Print version information and exit

License and citation

This research is released under the CC BY-NC 3.0 US license. We encourage an attitude of reproducible work for academic-only purpose. Please kindly cite our work in your publications if it helps your research:

 author = {Zang, Guangming and Idoughi, Ramzi and Tao, Ran and Lubineau, Gilles and Wonka, Peter and Heidrich, Wolfgang},
 title = {Space-time Tomography for Continuously Deforming Objects},
 journal = {ACM Trans. Graph.},
 issue_date = {August 2018},
 volume = {37},
 number = {4},
 month = jul,
 year = {2018},
 issn = {0730-0301},
 pages = {100:1--100:14},
 articleno = {100},
 numpages = {14},
 url = {},
 doi = {10.1145/3197517.3201298},
 acmid = {3201298},
 publisher = {ACM},
 address = {New York, NY, USA},
 keywords = {4D reconstruction, X-ray computed tomography, optimization},


If you find any bug or if you have any suggestion or comment, please contact:

Guangming :

Copyright: Visual computing center, CEMSE, KAUST



This is the open source repository for SIGGRAPH 2018 paper "Space-time Tomography for Continuously Deforming Objects".




No releases published


No packages published