z-logo
Premium
Calibrating Corn Color from Aerial Photographs to Predict Sidedress Nitrogen Need
Author(s) -
Scharf Peter C.,
Lory John A.
Publication year - 2002
Publication title -
agronomy journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.752
H-Index - 131
eISSN - 1435-0645
pISSN - 0002-1962
DOI - 10.2134/agronj2002.3970
Subject(s) - sowing , zea mays , nitrogen , agronomy , environmental science , crop , pixel , mathematics , chemistry , biology , artificial intelligence , computer science , organic chemistry
Supplemental N need of corn ( Zea mays L.) and other crops can vary substantially within and among fields. Corn color is sensitive to N status and may provide a means to accurately match N fertilizer rates to spatially variable N needs. Our objective was to calibrate the relationship between corn color measured in aerial photographs and sidedress N need. Economic optimum N rate (EONR) at sidedress was determined in 18 yield response experiments located in production cornfields. Low‐altitude, high‐resolution aerial photographs were taken at growth stage V6 or V7 with two types of film: color positive and color infrared. The EONR ranged from 0 to 336 kg N ha −1 . For both types of film, corn color was a significant predictor of EONR at sidedress but only when expressed relative to the color of well‐fertilized corn in the same field and when no N had been applied at planting. Predictions were more accurate using color film than color‐infrared film. Removal of soil pixels from the true‐color aerial images greatly strengthened the relationship between measured color and EONR: R 2 values ranged from 0.27 to 0.31 for single colors measured from the entire image and from 0.60 to 0.79 after the removal of soil pixels. Our results demonstrate that corn color measured in aerial photographs can be used to predict sidedress N need. Obstacles to practical use in guiding variable‐rate sidedressing include: no N can be applied at planting, a high‐N reference strip is needed, and soil pixels must be removed from the image.

This content is not available in your region!

Continue researching here.

Having issues? You can contact us here