Skip to content

First KNOSSOS Qt Release
Compare
Choose a tag to compare
@Optiligence Optiligence released this 03 Jun 23:12
· 3253 commits to master since this release
v4.0
f692c7d

KNOSSOS v4.0 - Change Log

KNOSSOS is a software tool for the visualization and annotation of 3D image data and was developed for the rapid reconstruction of neural morphology and connectivity.

General Information

Release Date: Date 03.06.2014

KNOSSOS 4.0 is a milestone release with a completely reworked graphical user interface based on
the QT framework.

Homepage: http://www.knossostool.org/
Project host: https://github.com/knossos-project/

Roadmap:
v4.1:

  • Interactive Python scripting
  • Volume annotation and visualization of large scale segmentation data

Known Bugs

UI Improvements

  • Nicer look and feel
  • Native file dialogs and ui elements
  • Restructured menu:
    • Icons and shortcut information in the menu bar
    • Current workmodes are marked with a check
    • A Toolbar with a collection of important widgets
    • Restructured Viewport Settings Widget
  • Tooltips
  • Integrated documentation
  • Unsaved changes are indicated with an (*) in the titlebar

Features

Tracing:

  • Integration of client-server based task management (with a Python django backend), that allows annotators to efficiently submit
    their tracings to a database.
  • Extensive filtering and listing of trees and nodes (with regular expression support) based on comments and other properties.
  • Visual selection and operations on groups of nodes in all viewports.
  • Show only selected trees.

Dataset:

  • Real time arbitrary angle slicing for voxel data (use eg l or k for rotations).
  • Datasets can now be streamed from remote server locations (http based, works well for connection speeds > 10 MBit/s with compression)
  • Datasets can be compressed (each datacube individually) with JPG2000 or JPEG with real time decompression during the navigation (multi-threaded).
    Typical values for 3D EM datasets without obvious compression artefacts are ~1 bit / voxel. Lossless compression is also supported.
  • On-the-fly dataset changes possible under File->Load Dataset, no more Python Launcher required.
  • KNOSSOS remembers the last used dataset and loads it on launching.

Skeleton:

  • Multiple skeletons can now be loaded at the same time (select more than one in the file dialog).
  • Load Skeletons by drag'n dropping them into KNOSSOS.

Changes

  • The Python Launcher has been removed.
  • Moving Viewports around now works with Ctrl+AltLeft mouse drag
  • The synchonization of multiple KNOSSOS instances is currently not supported any more
  • The format for custom preferences is now saved in a key value file structure (.conf)
  • Default Folder for Skeleton Files is now in the standard data location (i.e for Windows: %AppData%/MPIMF and for Linux: /home/user/.local/share/data/MPIMF)
  • A Mac OS X Version (>= version 10.9) will be available soon.
  • Debian Package for Linux (Tested on Ubuntu [i386/amd64](>= version 12))
Assets 6
Loading