Phenomenal: a software framework for model-assisted analysis of high throughput plant phenotyping data

摘要: Plant high-throughput phenotyping aims at capturing the genetic variability of plant response to environmental factors for thousands plants, hence identifying heritable traits genomic selection and predicting values allelic combinations in different environment. This first implies automation measurement a large number characterize growth, development functioning. It also requires fluent versatile interaction between data continuously evolving models, that are essential analysis marker x environment integration processes crop performance [1]. In frame Phenome high throughput infrastructure, we develop Phenomenal: software framework dedicated models. is based on OpenAlea platform [2] provides methods softwares modelling together with user-friendly interface design execution scientific workflows. part InfraPhenoGrid infrastructure allows computation recording provenance during [3]. Figure 1: The 3D reconstruction segmentation pipeline. Muti-view plants images from PhenoArch binarised used reconstruct in3D. skeleton extracted separated into stem (central vertical elements) leaves. voxels segmented by propagating segmentation.