See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/2606184 Presentation Outline Article · July 1999 Source: CiteSeer CITATIONS READS 0 16 2 authors , including: P. Khargonekar University of Florida 416 PUBLICATIONS 24,002 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Signal Processing via Sampled-data Control Theory View project Control & Combinatorial Optimization View project All content following this page was uploaded by P. Khargonekar on 26 March 2015. The user has requested enhancement of the downloaded file.
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. AVIRIS WORKSHOP 2000 Precision Farming Application of AVIRIS Data in the San Joaquin Valley, California Presented at the AVIRIS Earth Science and Applications Workshop February 23-25, 2000 JPL, PASADENA, CA WORK SPONSORED IN PART BY THE NASA/ SSC HYPERSPECTRAL EOCAP Nahum Gat 1 , Hector Erives 1 , Glenn J. Fitzgerald 2 , Stephan J. Maas 2 1 Opto-knowledge Systems, Inc. (OKSI) 2 USDA-ARS, Western Integrated Cropping Systems Research Unit, Shafter, CA In colaboration with Susan Ustin & the CSTARS Lab at UC Davis
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 PRESENTATION OUTLINE PROJECT OBJECTIVES APPROACH LAB WORK FIELD WORK (EXPERIMENTAL PLOTS) AIR BORNE (EXPERIMENTAL & COMMERCIAL PLOTS) PRELIMINARY RESULTS CONCLUSIONS (& FUTURE WORK)
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 PROJECT OBJECTIVES OVERALL OBJECTIVE: HELP INCORPORATE REMOTE SENSING INTO PRECISION FARMING PRACTICES EXPLORE THE USE OF HYPERSPECTRAL TECHNOLOGY IN REAL-TIME COMMERCIAL CROP MANAGEMENT SPECIFIC OBJECTIVE: EARLY DETECTION OF STRESS; IN PARTICULAR PEST INSECT INFESTATION EMPHASIS ON COTTON
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 LAB WORK Spectral analysis of samples Biomass Insect count
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 Laboratory spectral characterization Classification of lab Spectra of healthy & Healthy & infested samples infested cotton cotton samples samples
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 FIELD WORK Canopy-level (imaging) spectra Spectra library development (plants, soils, etc.)
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 Canopy-scale field characterization (using LCTF camera system) Scale of field characterization (580 Field hyperspectral close up nm image) characterization
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 GROUND TRUTHING FOR AIRBORNE IMAGERY
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 MET STATION AND IR RADIOMETERS & DATA LOGGER
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 AIRBORNE: SAMRSS AIRBORNE MULTI SPECTRAL SENSOR 3 CCD CAMERAS 1024×1024 PIXELS, 12 BITS BANDS SELECTION WITH FILTERS 548, 658, 850 nm FWHM: 13 nm FOR GREEN/RED, 40 nm FOR NIR THERMAL IR CAMERA (NEW) COLOR VIDEO CAMERA DATA ACQUISITION TO PC
SAMRSS MULTI-SPECTRAL SENSOR IMAGERY (GSD=0.6 m) Experimental plots; calibration panels for SAMRSS
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 CALIBRATION PANELS FOR SAMRSS
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 New 400-1000 nm pushbroom sensor for Spring 2000
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 COMMERCIAL SCALE PLOTS Sheely Farm
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 LOCATION: SAN JOAQUIN VALLEY, CA FRESNO 99 Fwy ~80 air- Sheely Farm miles to Dryden Lemoore Westlake Farm 5 Fwy
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 AVIRIS FLIGHT LINE
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 1999 CROP MAP TOMATO GARLIC PISTACHIO PIMA COTTON TOMATO COTTON COTTON GARBANZO COTTON COTTON TOMATO PIMA
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 Plot # Date Operation Com m e nts OPERATIONS SCHEDULE, 6-1 April 04/18/99 Planting Maxxa AVIRIS FLIGHTS, & CROP May 05/14/99 Cultivate 06/05/99 Cultivate SIGNATURES (COTTON EXAMPLE) June 06/05/99 Insecticide grnd 06/07/99 Insecticide grnd Flight 6/13/99 06/13/99 Herbicide grnd 06/14/99 Fertilize W ater run July 07/01/99 Cultivate 07/04/99 Insecticide grnd Plot: 6-1 (cotton) 07/04/99 Fertilize W ater run 07/22/99 Fertilize W ater run 7 06/13/99 6 08/02/99 Irrigation Furrow Reflectance x 10,000 08/28/99 08/04/99 Insecticide Aerial 5 09/01/99 08/19/99 Irrigation Furrow Thousands 4 Flight 8/28/99 09/10/99 08/31/99 Insecticide Aerial 3 09/24/99 Flight 9/1/99 2 Flight 9/10/99 Flight 9/24/94 1 10/02/99 Defoliant prep Aerial 0 10/09/99 Defoliant Aerial 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 10/20/99 Harvest Wavelength, µm
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 Plot # Date Operation Com m e nts OPERATIONS SCHEDULE, 6-2 April 04/18/99 Planting Phytogen AVIRIS FLIGHTS, & CROP May 05/12/99 Cultivate 05/20/99 Herbicide grnd SIGNATURES 05/20/99 Fertilizing Sidedress 05/27/99 Cultivate (PIMA EXAMPLE) June 06/09/99 Herbicide grnd Flight 6/13/99 06/13/99 Fertilize W ater run 06/30/99 Cultivate July 07/01/99 Insecticide grnd 07/02/99 Fertilize W ater run Region: 6-2 (PIMA) 07/15/99 Fertilize W ater run 07/27/99 Fertilize W ater run 7 07/28/99 Insecticide Aerial 06/13/99 6 Reflectance x 10,000 08/10/99 Irrigation Furrow 08/28/99 5 08/20/99 Irrigation Furrow 09/01/99 08/25/99 Irrigation Furrow Thousands 4 09/10/99 Flight 8/28/99 09/24/99 3 08/29/99 Insecticide Aerial Flight 9/1/99 2 Flight 9/10/99 1 Flight 9/24/94 10/09/99 Defoliant prep Aerial 0 10/18/99 Defoliant Aerial 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 11/04/99 Harvest Wavelength, µm 11/10/99 Harvest
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 7 SIGNATURES 6/13/99 Region: 4-1 (COTTON) 6 Reflectance x 10,000 Region: 5-1 (COTTON) 5 Region: 6-1 (COTTON) Region: 5-2 (PIMA) Thousands Emergence stage 4 Region: 5-3 (PIMA) Region: 6-2 (PIMA) 3 2 1 0 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 Wavelength, µm Pima vs. Cotton; plot-averaged, peak canopy development) Plot-averaged signatures (peak canopy development) 7 7 8/28/99 8/28/99 6 6 Reflectance x 10,000 Reflectance x 10,000 5 5 Thousands Thousands 4 4 3 Region: 4-1 (COTTON) 3 Region: 5-1 (COTTON) 2 Region: 6-1 (COTTON) 2 Region: 4-1 (COTTON) Region: 5-2 (PIMA) Region: 6-2 (PIMA) Region: 5-3 (PIMA) 1 1 Region: 6-2 (PIMA) 0 0 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 Wavelength, µm Wavelength, µm
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 SIGNATURE ANALYSES Review vegetation indices Photosynthetic activity indices Canopy water indices Review eigen-indices Remove redundancy Capture prime factors indication plant vigor / stress Review other parameters that affect plant health soil parameters, weather Consider that plant is an integrator of the conditions that affect health / stress Preliminary conclusions: 1. Need to work with time-integrated parameters 2. Need "many" spectral bands to capture all relevant factors affecting plant growth and reduce reliance on ground measurements 3. Quick results obtained with VNIR bands using Tasseled Cap type analysis † (see this morning paper on sugar beet yield prediction)
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 TIME-INTEGRATED INDICES TOMATO COTTON / PIMA (Some) Time-integrated indices are species-specific because of differences in growth cycle
OKSI - OPTO-KNOWLEDGE SYSTEMS, INC. Precision Farming (Gat, et-al ) AVIRIS WORKSHOP 2000 SOILS / CROP STATUS CLASSIFICATION Use AVIRIS imagery to develop a classification that classifies: Soils (dark, light) Vegetation (emerging, mature and senescing) Transformation based on 4 bands: 0.4588, 0.5571, 0.6553, 0.8163 (Gram-Schmidt orthogonalization) Use soil spectra to form a base normal vector Brightness axis Use vegetation spectra to form a second orthonormal axis vector Greenness Use senescing vegetation to form a third orthonormal axis vector Yellowness Apply transformation to imagery V e g e t a t i o n . g e v e r u t a M . g e v Greenness g n i g r e m E e n i l l "Tasseled Cap"-like approach i o S Soil Brightness
Recommend
More recommend
