Rhizoscan – file format

Standard root architecture file format:

An international collaboration between several of the main software developers of root measurement (from images) has emerged. A meeting has been arranged in order to define a standard file format for serialization and exchange of root architecture data.

With Guillaume Lobet (developer of SmartRoot), I have volunteer to organize this meeting which will occur the 18th to the 20th February either in UCL (Louvain-la-neuve) or in ULg (Liège).


(underline indicates the participant to the meeting)

RootReader 2D/3D (Cornell): Randy Clark, Leon V. Kochian

RootNav (CPIB): Michael Pound, Malcolm Bennett, Tony Pridmore, Stefan Mairhofer

SmartRoot (UCL&ULg): Guillaume LobetXavier Draye

RhizoScan (Virtual Plants): Julien DienerChristophe Pradal, Christophe Godin – Philippe Nacry (INRA)

Root System Analyser (UofVienna):  Daniel LeitnerAndrea Schnepf

  • List the meeting deliverable
  • Make the meeting planning
  • define a first (but valid) format to be used as starting model
  • Update AxialRootTree.to_mtg() to be compliant with this format (warning account for multi-axes per segment)
  • make a starting python package for this format
    • add a mtg loader
    • add simple measurements tool (from mtg)
    • add comparison tool from two mtg objects (to replace current RootAxialTree comparison)
  • Make a simple upload/download/file-format-validation server?


Thought about meeting preparation:


  • 1st, as soon as possible:
    • Exact date (and time) of meeting start and end
    • Guillaume and I will provide a planning beginning of february
  • 1st or 2nd?: ask to prepare a presentation highlighting
    • Current soft serialization method, and stored information (deciding if those are topological or geometrical, modules or properties or metadata
    • how they manage time, if done?
    • context and applicative objectives (measurements,


Information should be differentiated as topological or geometrical

In the storage format, properties can be structural (implied by the format structure) or “decoration”

the more precis and unique the format is the better, but thus impose more development constraint on each

Leave a Reply

Your email address will not be published. Required fields are marked *