the RoSoM project: root system analysis

Influence of root-soil mechanical interaction on the variability of root architecture (RoSoM)

Keywords — root system, root phenotyping, image processing, geometry and density characterization

In this project, we propose a new global approach to characterize the geometry of the root system form from black-and-white silhouettes. We introduce a covering polygon composed of two superimposed quadrangles to evaluate geometry features or to classify the growth strategy of root systems: the convex hull of the root system is progressively reduced to a double-quadrangle-shaped polygon by iterative suppression of least significant vertices to only keep the major geometry features.


Fig. 1: the double-quadrangle-shaped polygon and its masks. Different steps of the convex hull regression are presented on a mature and dense rice root system (30 days). From left to right, and top to bottom: the convex hull and the 9th, 15th, 16th, 17th, 18th step with respectively 28, 19, 13, 12, 11 and 9 vertices. The covering border is represented in red. The 1%-upper and 20%-lower masks (in orange) cover the zones where the vertices of the convex hull could not be removed, and the complementary zone – in green – where the convex hull reduction will be applied.

The root system density is addressed by the analysis of the hole distribution. The root system porosity is estimated from the decomposition of background regions in circular elements. An automated repartition of circular elements by 2-means clustering is introduced for a best appreciation of macro- and micro-holes, respectively associated to the anchorage and nutrient recovery potentials.

Fig. 2: “circular hole” distribution into the double-quadrangle-shaped polygon. The mature root is system is represented in green, the micro-holes in blue and the macro-holes in red. d- 4-week old wheat, e- 3-week-old sorghum, f- 6-week-old palm tree.

The global parameters of the root system are evaluated from the geometry and density properties of the double-quadrangle-shaped polygon. For more détails, see the FSPMA 2016 poster.


GT-RootS (Global Traits of Root System) is a automated Java-based open-source solution we develop for processing the root system images coming from the Rhizoscope, a Cirad’s phenotyping plateform dedicated to adult cereal plants.


Borianne, P., Subsol, G., Fallavier, F., Dardou, A., Audebert, A. (2018). GT-RootS: An integrated software for automated root system measurement from high-throughput phenotyping platform images. Computers and Electronics in Agriculture 150 (2018) 328–342

Borianne, P. ; Subsol, G., Audebert, A. (2016). Automated Characterization of the Mature Root System Form by a Double-Quadrangle-Shaped Polygon. IEEE International Conference on Functional-Structural Plant Growth Modeling, Simulation, Visualization and Applications (FSPMA), Qingdao, China, pp. 6-15. doi: 10.1109/FSPMA.2016.7818282.




This work is supported by Agropolis Fondation under the reference ID 1202-073 through the “Investissements d’avenir” program (Labex Agro: ANR-10-LABX-001-01).

Contact :

Image aspects, Phenotyping aspects




L'image au service de l'étude des Plantes et des Paysages