Vision Based Modeling of Plants Phenotyping in Vertical Farming under Artificial Lighting

Abstract

In this paper, we present a novel method for vision based plants phenotyping in indoor vertical farming under artificial lighting. The method combines 3D plants modeling and deep segmentation of the higher leaves, during a period of 25–30 days, related to their growth. The novelty of our approach is in providing 3D reconstruction, leaf segmentation, geometric surface modeling, and deep network estimation for weight prediction to effectively measure plant growth, under three relevant phenotype features: height, weight and leaf area. Together with the vision based measurements, to verify the soundness of our proposed method, we also harvested the plants at specific time periods to take manual measurements, collecting a great amount of data. In particular, we manually collected 2592 data points related to the plant phenotype and 1728 images of the plants. This allowed us to show with a good number of experiments that the vision based methods ensure a quite accurate prediction of the considered features, providing a way to predict plant behavior, under specific conditions, without any need to resort to human measurements.

Plant phenotyping relevance

植物フェノタイピングのための3D再構成・葉セグメンテーション・深層学習による形質推定法の開発と手測定による検証が研究の中心である。

Code and data availability

The paper's leaf segmentation analysis code (a Tensorflow Mask R-CNN implementation used for the plant phenotyping measurements) is explicitly stated to be freely available on the authors' GitHub organization. The collected image/phenotype dataset (1728 images, 2592 manual data points) is described but no public data-­