Commits

Branch selector

 main

User selector

All users

All time

Commit history

Commits on Jun 18, 2026

• cleanup + sketch of a windows installer

  sporring

  committed
  3392aa3

  Copy full SHA for 3392aa3

Commits on Jun 17, 2026

• ubuntu stuff

  sporring

  committed
  b1a463a

  Copy full SHA for b1a463a

• removed all traces off SharpFFTW

  sporring

  committed
  e8c6a8b

  Copy full SHA for e8c6a8b

• still depending on the Zarr.NET fork

  sporring

  committed
  892527c

  Copy full SHA for 892527c

• removed leftover simple-itk and legacy code

  sporring

  committed
  66582fb

  Copy full SHA for 66582fb

• Merge pull request #14 from sporring/simd-slices

  Show description for fb0e1a4
  sporring

  authored
  fb0e1a4

  Copy full SHA for fb0e1a4

• multipage tiff read is slow. working on it, but want pull all of this into the main branch.

  sporring

  committed
  3f42b42

  Copy full SHA for 3f42b42

• extracted Image - the simple-itk wrapper - to a repository on its own

  sporring

  committed
  b87698e

  Copy full SHA for b87698e

• ran zarr-fast benchmarks.

  sporring

  committed
  4af06bc

  Copy full SHA for 4af06bc

Commits on Jun 16, 2026

• studio error being chased

  sporring

  committed
  58c1546

  Copy full SHA for 58c1546

• rerun all lmip benchmarks for pdf figures. working on zarr benchmark. Updated sp-zarr reader and writer

  sporring

  committed
  9dd86fe

  Copy full SHA for 9dd86fe

Commits on Jun 15, 2026

• minor bug

  sporring

  committed
  a1519a2

  Copy full SHA for a1519a2

• cleanup notes

  sporring

  committed
  f724c7c

  Copy full SHA for f724c7c

Commits on Jun 14, 2026

• cleaning up

  sporring

  committed
  751116b

  Copy full SHA for 751116b

• made a lean tiff read and writer, speed now much better. fftw xy+z stage no seems to work, and uses a compact and fast real->complex->real path

  sporring

  committed
  c65d76e

  Copy full SHA for c65d76e

Commits on Jun 13, 2026

• updated convolve

  sporring

  committed
  44b1824

  Copy full SHA for 44b1824

Commits on Jun 12, 2026

• renamed user facing functions to hide irrelevant implementation details. warped benchmarks to remove old image stages

  sporring

  committed
  6f50bc0

  Copy full SHA for 6f50bc0

• Image has been flushed

  sporring

  committed
  a2c7702

  Copy full SHA for a2c7702

• everything should be Chunked now. Getting ready to disconnect Image entirely.

  sporring

  committed
  6479d17

  Copy full SHA for 6479d17

• almost all samples are now Chunked, I think it's about time to let go of Image and Slab

  sporring

  committed
  b714cdb

  Copy full SHA for b714cdb

• decided to shelve gray-value morphology. Now there are only few samples, which still needs wiring.

  sporring

  committed
  cd375c4

  Copy full SHA for cd375c4

• more studio/Image ported to Chunk

  sporring

  committed
  2168c14

  Copy full SHA for 2168c14

• Going for the full replacement of Image by Chunk: Added many missing pieces

  sporring

  committed
  64ee852

  Copy full SHA for 64ee852

Commits on Jun 11, 2026

• all major algorithms implemented. affine resample, fft, signed distance ban have not been optimised yet.

  sporring

  committed
  5609302

  Copy full SHA for 5609302

• ported affineResample, Bias correct, added padding, concatAlong, crop, and transpose functions

  sporring

  committed
  740e50b

  Copy full SHA for 740e50b

• added fast 1d axis aligned convolution. Cleaned up Chunk processing by adding a Chunk pure project and separating all the hard algorithms in Core. Added separable filters.

  sporring

  committed
  2094cd9

  Copy full SHA for 2094cd9

Commits on Jun 9, 2026

• working on fft. Presently zarr read-write takes all the time, and it's much faster for copying, so something is wrong.

  sporring

  committed
  eea75c4

  Copy full SHA for eea75c4

• cunk fft added. full vollume c++-itk takes 0.415, sp-2-pass takes 1.219, which is fine, since it has an extra write-read zarr image part.

  sporring

  committed
  ec4d4c0

  Copy full SHA for ec4d4c0

• first version of chunk-fftw

  sporring

  committed
  c6851d5

  Copy full SHA for c6851d5

• working with remaining tasks

  sporring

  committed
  afe9b96

  Copy full SHA for afe9b96

Commits on Jun 8, 2026

• In-house connected components obtained. Implemented a faster convolve backend, now very close to c++-itk. Also, started looking at releaseTrimmed and AOT optimisation of dotnet code.

  sporring

  committed
  0a3f82b

  Copy full SHA for 0a3f82b

• median added for the other typical types, int32 raised as part of the gang as the signed work horse.

  sporring

  committed
  c314d09

  Copy full SHA for c314d09

• wrapping c++ nth_element for most medians seems to be the only way

  sporring

  committed
  f9e44bc

  Copy full SHA for f9e44bc

• median uint8 experiements run, more single-pixel operations moved to chunk

  sporring

  committed
  868e446

  Copy full SHA for 868e446

Commits on Jun 7, 2026

• a working parallelised ph for up to 39^3 neighbourhoods.

  sporring

  committed
  bb63014

  Copy full SHA for bb63014