Database Research Summaries2018 ND Identify and Develop Glyphosate Resistant Weed Maps in Soybean Fields
Research Focus
The focus of this project is to use digital RGB, multispectral and thermal imaging as a remote sensing technique to detect and accurately identify weeds.
Objectives
- Field test and validate a new methodology developed at NDSU to identify herbicide resistant weeds and weed species in commercial soybean fields using UAS-based imagery.
- Develop a GIS platform for sharing weed map data of commercial fields with farmers.
Results
- There is a significant difference between spectral reflectance of the three weed species at specific wavelengths. Soft Independent Modeling of Class Analogy (SIMCA) analyzed spectral reflectance of weeds.
- The model indicated that red, red-edge region (640, 676, and 730 nm) had the best wavelengths for weed discrimination in visible range. Multispectral sensor may have potential to identify weed species in the field. Multispectral images were analyzed to find the best supervised or unsupervised method for weed species discrimination based on multispectral imagery.
- The plant canopy temperature extracted from the thermal images indicated that the temperature of susceptible weeds increased soon after spraying glyphosate.
- The support vector machine (SVM) method used canopy temperature to classify resistant and susceptible weeds with 100 % accuracy before any visual symptoms were apparent.
- Results showed that 15, 19, 28, 36, 44, 46, 60, 70 and 91 hours after herbicide application were the best windows for identifying glyphosate-resistant weeds from susceptible ones.
Importance
Remote sensing will provide an inexpensive and more efficient method for mapping weed infestations than ground surveys. Early and easier detection of herbicide resistant weeds would allow growers to effectively manage these weeds by applying another herbicide, or using manual method before the plant seeds and spreads.
For more information about this research project, please visit the National Soybean Checkoff Research Database.
Funded in part by the soybean checkoff.
