Microwave backscatter response of pecan tree canopy samples for estimation of pecan yield in situ using terrestrial radar

  • Authors:
  • James A. Hardin;Paul R. Weckler;Carol L. Jones

  • Affiliations:
  • Department of Biosystems & Agricultural Engineering, Oklahoma State University, 111 Ag Hall, Stillwater, OK 74078, United States;Department of Biosystems & Agricultural Engineering, Oklahoma State University, 111 Ag Hall, Stillwater, OK 74078, United States;Department of Biosystems & Agricultural Engineering, Oklahoma State University, 111 Ag Hall, Stillwater, OK 74078, United States

  • Venue:
  • Computers and Electronics in Agriculture
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Accurate estimates of pecan yield prior to harvest are critically important for marketing and production management decisions such as nut thinning, irrigating and nutrition supplementing. Current methods of estimating pecan yield in situ are not sufficiently accurate and are time consuming. Research using satellite based microwave imaging has enabled scientists to identify trends in orchard crop condition but precision is inadequate for yield sensing. Ground based radar schemes using antenna within the orchard resolve many of the power, resolution and sensitivity limitations of satellite radar imagery. The objective of this research is to determine if pecan nuts can be quantified in situ using backscattered microwaves from antenna located in the orchard. Pecan tree canopy samples (leaves and secondary branches) and nuts were collected at five growth stages and placed in a polystyrene foam test fixture located between horn antennae spaced 1m apart. Reflection and transmission measurements were recorded with a vector network analyzer at frequencies from 1 to 18GHz while the amount of nuts were varied from 0% to approximately 30% of the canopy mass. Regression analysis revealed no specific frequencies to quantify nut mass however response to total canopy water and dry mass over a wide range of frequencies had R^20.63 and 0.78 respectively. This relationship combined with range finding and appropriate crop model algorithms may ultimately be the basis for developing pecan yield monitoring technology.