Weed species differentiation using spectral reflectance land image classification

J. T. Sanders; W. J. Everman; R. Austin; G. T. Roberson; R. J. Richardson

doi:10.1117/12.2519306

10 May 2019 Weed species differentiation using spectral reflectance land image classification

J. T. Sanders, W. J. Everman, R. Austin, G. T. Roberson, R. J. Richardson

Proceedings Volume 11007, Advanced Environmental, Chemical, and Biological Sensing Technologies XV; 110070P (2019) https://doi.org/10.1117/12.2519306
Event: SPIE Defense + Commercial Sensing, 2019, Baltimore, MD, United States

Abstract

Advancements in efficient unmanned aerial platforms and affordable sensors has led to renewed interest in remote sensing by agricultural producers and land managers for use as an efficient and convenient method of evaluating crop status and pest issues in their fields. For remote sensing to be employed as a viable and widespread tool for weed management, the accurate detection of distinct weed species must be possible through the use of analytical procedures on the resultant imagery. Additionally, the remote sensing platform and subsequent analysis must be capable of identifying these species across a wide range of heights. In 2017, a field study was performed to identify any weed height thresholds on the accurate detection and subsequent classification of three common broadleaf weed species in the southeastern United States: Palmer amaranth (Amaranthus palmeri), common ragweed (Ambrosia artemisiifolia) and sicklepod Senna obtusifolia) as well as the classification accuracy of image classifications performed on the species scale. Pots of the three species at heights of 5, 10, 15, and 30 cm were randomly arranged in a grid and 5-band multispectral imagery was collected at 15 m. Image analysis was used to identify the spectral reflectance behavior of the weed species and height combinations and to evaluate the accuracy of species based supervised classifications involving the three species. Supervised classification was able to discriminate between the three weed species with between 24-100% accuracy depending on height and species. Palmer amaranth classification accuracy was consistently 100%. Increased height of sicklepod and common ragweed plants did not reliably confer improved accuracy but the species were correctly identified with at least 24% and 60% accuracy, respectively.

Citation Download Citation

J. T. Sanders, W. J. Everman, R. Austin, G. T. Roberson, and R. J. Richardson "Weed species differentiation using spectral reflectance land image classification", Proc. SPIE 11007, Advanced Environmental, Chemical, and Biological Sensing Technologies XV, 110070P (10 May 2019); https://doi.org/10.1117/12.2519306

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available