Enterprise Machine Intelligence and Learning Initiative

Published purpose

Implementation of digital agriculture through high throughput-imaging is key to improving plant phenotyping for crop improvement. Image analyses utilizing machine learning (ML) approaches have the potential to dramatically improve crop phenotyping. The intent of the project is to capture above-ground image datasets of peas in an agricultural field, establish an ML pipeline to identify plants and quantify biomass, and begin to associate field traits such as yield and protein with beneficial root traits. For roots, an automated ML pipeline will be established to characterize rhizobium nodules and root system architecture in controlled environments. The foundational datasets generated will enable field to lab comparisons and importantly facilitate the development of advanced tools deployable for improving pea production in the Canadian Prairies