Large scale phenotyping can help us understand biology but also to adapt cultures to a changing 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. There is therefore a need of automated procedures to quickly extract important parameters from live models. In this work, we propose a fully automated procedure to extract the structure of arabidopsis, including internode lengths and angles be- tween successive organs. In arabidopsis like in many other plants, the sequence of organs produce a Fibonacci spiral, each organ having an angle around the golden angle of 137.5◦ to the previous organ. Anomalies in these sequences are of particular interest to biologists. The paper is structured as follows: section 2 reviews the related literature. Section 3 presents the general methodology and the hardware setup used in the experiments. Section 4 describes into details the different steps of the reconstruction procedure. Finally, section 5 validates the approach on both simulated and real world data.
