Theme I2P - Imaging for Plants and landscaPes

Context

Using numerical images as supports for observations, restitutions and measurements in plant biology and ecology is becoming a common practice with increasing access to systems of numerical image acquisition or to the use of realistic simulation models: microscopy imaging to understand the internal development of wood, aerial images for the study of gaseous exchanges between plants and the atmosphere, 3D reconstruction to understand plant / environment interactions, etc. If the nature and the content of the images depend on the scientific question, their processing to identify and characterize visible homogeneous structures is generic. The automation of these processes enables large-scale analyses to identify trends and invariants.

Scientific objective

The group aims to adapt well-known methods of image processing or image synthesis, or to develop new approaches to plants and landscapes to facilitate the use of numerical contents for applied studies, especially in landscape ecology, plant or complex agro-forestry system diagnosis, the production or calibration of growth models. The group also performs expertise in imaging and provides results and/or dedicated software tools to enable botanists or ecologists to answer their own scientific questions.

Approach

The group promotes automated (or at least auto settable) approaches (of analysis or learning) in image processing or image synthesis. The challenge is to develop new human-operator-independent methods that ensure the reproducibility and the quality of the results according to the evaluated confidence intervals. Concerning image processing, some aspects of a given method may require further clarification according to the nature of the different images and the context in which they are acquired. For example, the group shares its experience in adapting the watershed image processing method to colored cells in microscope preparations or to hyperspectral remote sensing images. Concerning image synthesis, the group develops modeling approaches including both 3D dynamic representation and geometric (or spectral) characterization of the objects of interest (cells, whole plants, landscapes,etc.). The group is open minded, shares knowledge, experience and the tools needed to transpose and adapt efficient methods and algorithms from one application to another. Image processing, geometric modeling and image synthesis are intentionally associated to go beyond simple analysis towards a (computer-aided) diagnosis.

Expected results

  •  Sharing collective competence and experience whatever the nature and the scale of the processed images;
  •  Sharing codes and computer programs to more quickly identify and solve scientific problems; our tools are developed in an open-source environment to facilitate their testing, evaluation and diffusion;
  •  The production of numeric results or/and dedicated software tools in applicative projects involving the group (or some of the members or partners);
  •  The active participation in national or international scientific networks centered on the images.

Scientific projects

Acronym Title Duration
PIR-BCProche Infra Rouge Bas Coût : atelier scientifique Numev sur l’analyse spectrale des plantes2014 - 2015
TAFERApport des traits végétaux pour la compréhension et la visualisation des trajectoires successionnelles des communautés végétales de talus d’infrastructures linéaires de transport 2013 - 2016

Major publications

All publications
  • Borianne, P. ; Subsol, G., 2014. Fast Semi-supervised Segmentation of in Situ Tree Color Images. In : International Conference on Image and Signal Processing, Cherbourg (France), Lecture Notes in Computer Science 8509, Springer, p. 161–172.
  • Viennois, G.; Nicolini E.; Borne, F., 2014. Forest vs Savanna dynamics in the contact zones of French Guiana. Revue Française de Photogrammétrie et de Télédétection 208 [Editor link]
  • Brunel, G.; Borianne, P.; Subsol, G.; Jaeger, M., 2013. Simple-graphs fusion in image mosaic. Application to automated cell files identification in wood slices. In Kämäräinen, J. K., Koskela, M. (Eds). Image analysis : 18th Scandinavian Conference, SCIA 2013, Espoo, Finland, June 17-20, 2013. Proceedings. Berlin : Springer Verlag, p. 34-43 (Lecture notes in computer science).
  • Viennois, G.; Barbier, N.; Fabre, I.; Couteron, P., 2013. Multiresolution quantification of deciduouness in West central African Forest. Biogeosciences, 10(4) : 6957-6967. [Editor link]
  • Jaeger, M.,2012. Enhacing virtual natural scenes using quick and dirty image based recipes. In Kang, M. Z., Dumont, Y., Guo, Y. (Eds). Plant growth modeling, simulation, visualization and applications. Proceedings PMA12 : The Fourth International Symposium on Plant Growth Modeling, Simulation, Visualization and Applications, Shanghai, China, 31 October-3 November 2012. Beijing : IEEE Press, p. 212-219

Image gallery