University of Saskatchewan

Published purpose

The improvement and optimization of crop roots are a significant opportunity for ensuring crops are adapted, resilient and productive within Canada’s diverse hydro-climatic conditions. High-throughput phenotyping methodologies have been developed for crop roots but many challenges remain to optimize the imaging systems for automation efficiency, precision, and ease-of-use. Three dimensional (3D) imaging of plants is superior to two dimensional (2D) approaches and is being applied in numerous phenotyping applications globally to advance genetic gain. The Project focuses on the development of an enhanced 3D root phenotyping platform with step-change improvements to both imaging and trait analysis pipeline including data capture, segmentation, skeletonization, 3D reconstruction and trait extraction and analysis using AI and deep-learning methods as critical advancements. The ultimate output will be a high-throughput platform that facilitates superior 3D root phenotyping capturing low- and high-level vision details to quantitatively extract and model the complexity of crop root system architecture and development.