An automated image-processing pipeline

 for high-throughput analysis of root architecture in OpenAlea


click the image to see the publications of the RhizoScan project

Automated acquisition systems of Root System Architecture (RSA) are now readily available for developmental research and provide high-throughput image data of roots. Existing acquisition systems provide many types of data, from images of dispersed root pieces to full 3d scans of underground root systems. In this project we consider RSA grown in Petri plates. It is a traditional experimental protocol for which image data can be acquired easily and in large amount. Their analysis is thus a major challenge for researches on root development.

Existing tools range from the automatic estimation of global traits distribution to the extraction of more detailed representations of the root structure. To extract the full Root System Architecture, available softwares are either entirely manual or semi-automated. Those are suitable for detail inspection of architectural trait of a few root systems. But because of the time required by user interaction they cannot be used for high-throughput analysis.

In this project, we explore the possibility of extracting the full root architecture automatically and on a large-scale data set. Our framework is included in the OpenAlea platform which, in addition to traditional data analysis such as the ones provided by existing software, seamlessly provide the specialized architectural analysis packages already contained in OpenAlea.


  • Agropolis fondation with 24 month of postdoc in the Rhizopolis (21 month) and OpenAlea project (3 month)
  • Numev with 12 month of postdoc


  • Virtual Plants (INRIA, CIRAD, INRA) : J. Diener, F. Boudon, C. Pradal, C. Godin
  • BPMP (INRA) : P. Nacry, A. Gojon, H. Sentenac
  • Lepse (INRA) : B. Muller
  • DAP (CIRAD) : C. Périn, A. Dievart
  • UCL : X. Draye
  • Zenith (INRIA) : A. Joly, F. Masseglia

root_image&rsa w


  1. Pingback: Postdoc at Virtual Plants, INRIA | Julien.Diener

  2. Pingback: Julien Diener

Leave a Reply

Your email address will not be published.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