AUTHOR=Sanders John T. , Jones Eric A. L. , Minter Aiden , Austin Robert , Roberson Gary T. , Richardson Robert J. , Everman Wesley J. TITLE=Remote Sensing for Italian Ryegrass [Lolium perenne L. ssp. multiflorum (Lam.) Husnot] Detection in Winter Wheat (Triticum aestivum L.) JOURNAL=Frontiers in Agronomy VOLUME=3 YEAR=2021 URL=https://www.frontiersin.org/journals/agronomy/articles/10.3389/fagro.2021.687112 DOI=10.3389/fagro.2021.687112 ISSN=2673-3218 ABSTRACT=

Italian ryegrass [Lolium perenne L. ssp. multiflorum (Lam.) Husnot] is one of the most challenging weeds for winter wheat (Triticum aestivum L.) growers to manage. Italian ryegrass has evolved resistance to the majority of the herbicides labeled for use in wheat and the competitive ability of the species makes it a significant factor driving winter wheat production practices around the world. Previous research has utilized remotely sensed spectral imagery to detect Italian ryegrass in winter wheat to aid weed control decisions. Two studies from 2016 to 2017 were initiated with the intent of identifying the spectral reflectance properties of Italian ryegrass and winter wheat using an unmanned aerial vehicle (UAV) equipped with a 5-band multispectral sensor. Image analysis was conducted to determine the potential for species discrimination throughout the growing season. Supervised classification of the imagery was used to evaluate the ability of the UAV platform for further discrimination between Italian ryegrass and winter wheat. Species differentiation proved to be possible, however the data was not able to be referenced across dates. Due to light variability, the reflectance values changed to such a degree that unsupervised classifications were not possible using a database of values from previous flights. Supervised classification of the multispectral image resulted in >70% classification accuracy between the species. However, near infrared light consistently differed enough for accurate classification between Italian ryegrass and winter wheat across different weed densities, flight altitudes, and imaging dates. On a single field basis, species differentiation was successful and resulted in classified maps of Italian ryegrass and winter wheat. This study also analyzed the exact accuracy of the species differentiation based on the quality and uniformity of light conditions and growth stage of plants.