18th Symposium on Precision Agriculture in Australasia Wagga Wagga, 7th and 8th September 2015
CSIRO AGRICULTURE, HIGH RESOLUTION PLANT PHENOMICS CENTRE
Jose Jimenez-Berni, David Deery
LiDAR, thermal and hyperspectral sensors for crop monitoring applicaMons in PA
The need for a new green revoluMon
Source: Fischer et al, 2014. Crop yields and global food security: will yield increase continue to feed the world?
World yield of wheat, rice and maize. Average farm yield of wheat in Australia.
What is phenomics?
Genotype × Environment x Management
Weather Soil Ecosystem …
Irrigation Fertilizer Pesticides …
Variety
What is phenomics?
Genotype × Environment x Management = Phenotype
Yield Grain quality Disease resistance Canopy height Water use efficiency …
Weather Soil Ecosystem …
Irrigation Fertilizer Pesticides …
Variety
Phenomics from the pot to the farm
Thermal Lidar /Colour 3-‐D Chl Fluorescence Hyperspectral High spa>al or temporal resolu>on
The challenge of high throughput field phenotyping
Traits are dynamic!
Pt = G×Et (×M) with t: >me
Time constant
Seconds Hours Days Season
Canopy Temperature
Canopy Height
Crop Yield
Canopy temperature. The old way
h2=0.1
Canopy temperature. The new way
h2=0.7 Only a few seconds to cover the full trial
Airborne mosaic – mapping for spa>al variability
Yanco, NSW, early Oct 12:00, 350 m alt.
Extrac>on of plot temperature
My UAV Mmeline
2007
2008
2009
2010 2011
2012
2006
Our current fleet of Manned Aerial System
Phenomobile
• 3x LiDARs (Canopy Structure) • 4x RGB cameras (Stereo reconstrucIon)
• 1x Thermal IR camera (Canopy temperature)
• 1x Hyperspectral line scanner (Canopy biochemistry)
• 1x Full range spectrometer (Canopy biochemistry)
Deery et al, 2014
Phenomobile Lite
LiDAR • Canopy height • Fractional ground cover • Biomass index • Greenness vertical distribution GreenSeeker • NDVI (canopy density and
greenness) High-res RGB • Plant counts • Visual assessments
Why is LiDAR so cool?
Extracting information from the LiDAR
Soil elevation model Canopy elevation model
LiDAR Image
Spike counts from LiDAR
LiDAR is height and reflectance (active sensor)
LiDAR Intensity Height image
Comparison of fCover from LiDAR and NDVI (GreenSeeker)
Yanco BUNYIP 2014 (Tony Condon)
Preliminary validaMon of LiDAR: canopy biomass (wheat)
Sampling date 9/9/2014
Berni et al (in prep)
Processing Pipeline -‐ web interface
Processing Pipeline -‐ web interface
Micro-‐Hyperspectral imaging -‐ wet chemistry alterna>ve???
Leaf level predictions
Viridiana Silva, PhD Thesis
ApplicaMon to other crops and management
Precision irrigation
Weed management
Other horticultural crops
Acknowledgements
Xavier Sirault (Agricutlure flagship) Robert Furbank (Agriculture & ANU)
Jurgen Fripp (DP&S flagship) Helen Daily (Agriculture)
Peter Kuffner (Agriculture) Peter Ansell (Agriculture)
Julio Hernandez-Zaragoza (Agriculture) Dac Nguyen (Agriculture)
Robert Coe (IRRI) Chuong Nguyen (ANU)
Anthony Paproki (DP&S) Anne Bernhart (TelecomParis)
David Deery (Agriculture) Michael Salim (Agriculture) Jamie Scarrow (Agriculture)
Paul Hutchinson (Agriculture)
(lots of students…)
Christophe Pradal (INRIA, CIRAD) Christian Fournier (INRA)
Christophe Godin (INRIA, CIRAD) Francois Tardieu (INRA)
Frederic Baret (INRA) Pascal Neveu (INRA)
Paul Quick (IRRI)
Xinguang Zhu (Plant Systems Biology / CAS) Justin Borevitz (ANU)
Hamlyn Jones (University of Dundee)
Thank you AGRICULTURE FLAGSHIP/HIGH RESOLUTION PLANT PHENOMICS CENTER
For more info contact: [email protected]