WINTZ, COLLIAUX, HANAPPE: EXTRACTION OF TRAITS FROM ARABIDOPSIS 1 Automated extraction of phyllotactic traits from Arabidopsis thaliana Timothée Wintz wintz@csl.sony.fr David Colliaux colliaux@csl.sony.fr Peter Hanappe hanappe@csl.sony.fr Sony CSL Paris 6, rue Amyot 75005 Paris α l RGB 2D Segmentation Structure from Visual Hull 3d segmentation Angles Internode length Motion Sequence Figure 1: (Top left) Sample RGB image. (Top right) Measures of interest: angle between successive organs α ; internode length l . (Bottom) Proposed processing pipeline. 1 Introduction Large-scale phenotyping can help us understand biology but also to adapt cultures to a chang- ing environment. Working on well understood model plants like arabidopsis can both provide useful data to the biologists and is an important stepping stone to the study of more complex models: in-field imaging of crops, imaging of plant populations. . . Data collection and traits extractions are time-consuming process and a major bottleneck in plant phenotyping [13]. There is therefore a need of automated procedures to quickly c  2018. The copyright of this document resides with its authors. It may be distributed unchanged freely in print or electronic forms.