Plant high-throughput phenotyping using photogrammetry and imaging techniques to measure leaf length and rosette area

Abstract

Plant phenotyping is central to understand causal effects of genotypes and environments on trait expression and is a critical factor in expediting plant breeding. Previously, plant phenotypic traits were quantified using invasive, time-consuming, labor-intensive, cost-inefficient, and often destructive manual sampling methods that were also prone to observer error. In recent years, photogrammetry and image processing techniques have been introduced to plant phenotyping, but cost efficiency issues remain when combining these two techniques within large-scale plant phenotyping studies. Using these high-throughput techniques in basic plant biology research and agriculture are still in the developmental stages but show great promise for rapid phenotyping, which will materially aid both science and crop improvement efforts.In this study, we introduce an automated high-throughput phenotyping pipeline using affordable imaging systems and image processing algorithms to build 2D mosaicked orthophotos. Chamber-based and ground-level field implementations are used to measure phenotypic traits such as leaf length and rosette area in 2D images. Our automated pipeline has cross-platform capabilities and a degree of instrument independence, making it suitable for various situations.