Drone Phenotyping and Machine Learning Enable Discovery

Explore drone phenotyping and machine learning techniques for discovering loci regulating daily floral opening in lettuce in this webinar presented by Rongkui Han, a PhD student from the University of California. Delve into the innovative approach of using drones for field phenotyping and learn about image processing methods to analyze floral pixels. Understand the application of Support Vector Machine models and Bayesian statistics in this research. Examine the results showing high correspondence between blocks, phenotypic distribution, and genetic mapping. Investigate the detected QTLs and the relationship between floral opening time and flowering time. Gain insights into extracting floral pixels using machine learning and its implications for plant biology research.

Intro
KNOW YOUR LETTUCE
LETTUCE BELONGS TO THE COMPOSITEAE ASTERACEAE FAMILY
A "FLOWER CLOCK"
THE LETTUCE "FLOWER CLOCK": VISIBLE IN THE FIELD
SO WE DESIGNED AN EXPERIMENT
DRONE PHENOTYPING
DRONE IMAGES
IMAGE PROCESSING
THREE KINDS OF PIXELS
DEPLOYING SUPPORT VECTOR MACHINE MODEL
CHANGE OF FLORAL PIXELS OVER TIME
YES! USING BAYESIAN STATISTICS.
RESULTS: HIGH CORRESPONDENCE BETWEEN BLOCKS
PHENOTYPIC DISTRIBUTION
GENETIC MAPPING
QTL DETAILS
FLORAL OPENING TIME VS. FLOWERING TIME?
EXTRACTING FLORAL PIXELS USING MACHINE LEARNING