PhytoOracle: A Case Study in Automating Phenotyping
December 16, 2020 | Online
10:30AM–12:00PM US Central Standard Time
This team will detail the technologies, including CyVerse, and challenges of collecting and processing large phenotypic datasets and review the data pipeline, from collection to transfer to processing.
Transforming the Way We See Plants
As phenomics increasingly generates higher dimensional datasets, an urgent need to develop and implement robust data processing pipelines is emerging. The University of Arizona is home to the world's largest agricultural robot equipped with RGB, thermal, hyperspectral, and chlorophyll fluorescence cameras as well as a laser line scanner. PhytoOracle is being developed under AG2PI to address two key challenges encountered by phenomics research: (1) efficient, large-scale image stitching algorithm for high fidelity downstream analysis; (2) accelerating analyses by leveraging distributed computing resources. PhytoOracle's modularity handles increasing volumes and modalities of data making it easily customizable and scalable. As a result, PhytoOracle efficiently processes data to extract morphological and physiological parameters over a growing season. This presentation will focus on applications in plant research but will also include parallels to livestock research.
Upon registration, you will receive a confirmation email with information about joining the meeting.
A recording will be available at ag2pi.org at a later date.