GIS-based volunteer cotton habitat prediction and plant-level detection with UAV remote sensing

Abstract

Volunteer cotton plants germinate and grow at unwanted locations like transport routes and can serve as hosts for a harmful cotton pests called cotton boll weevils. The main objective of this study was to develop a geographic information system (GIS) framework to efficiently locate volunteer cotton plants in the cotton production regions in southern Texas, thus reducing time and economic cost for their removal. A GIS network analysis tool was applied to estimate the most likely routes for cotton transportation, and a GIS model was created to identify and visualize potential areas of volunteer cotton growth. The GIS model indicated that, of the 31 counties in southern Texas that may have habitat for volunteer cotton, Hidalgo, Cameron, Nueces, and San Patricio are the counties at the greatest risk. Moreover, a method based on unmanned aerial vehicle (UAV) remote sensing was proposed to detect the precise locations of volunteer cotton plants in potential areas for their subsequent removal. In this study, a UAV was used to scan limited samples of potential volunteer cotton growth areas identified with the GIS model. The results indicated that UAV remote sensing coupled with the proposed image analysis methods could accurately identify the precise locations of volunteer cotton and could potentially assist in the elimination of volunteer cotton along transport routes.