A model-based approach to recovering the structure of a plant from images

TL;DR

This paper introduces a silhouette-based, model-driven method for reconstructing detailed plant structures from a few images, focusing on phenotypic accuracy rather than just geometric shape, demonstrated on wheat.

Contribution

It presents a novel generate-and-test approach that uses a database of leaf models to accurately recover plant structure from minimal, widely spaced images without manual input.

Findings

Accurately reconstructs plant structure from few images

Applicable to various plant types, demonstrated on wheat

No manual intervention needed

Abstract

We present a method for recovering the structure of a plant directly from a small set of widely-spaced images. Structure recovery is more complex than shape estimation, but the resulting structure estimate is more closely related to phenotype than is a 3D geometric model. The method we propose is applicable to a wide variety of plants, but is demonstrated on wheat. Wheat is made up of thin elements with few identifiable features, making it difficult to analyse using standard feature matching techniques. Our method instead analyses the structure of plants using only their silhouettes. We employ a generate-and-test method, using a database of manually modelled leaves and a model for their composition to synthesise plausible plant structures which are evaluated against the images. The method is capable of efficiently recovering accurate estimates of plant structure in a wide variety of imaging scenarios, with no manual intervention.