Skip to main content
ARS Home » Plains Area » College Station, Texas » Southern Plains Agricultural Research Center » Aerial Application Technology Research » Research » Publications at this Location » Publication #333777
Title: Early identification of cotton fields using mosaicked aerial multispectral imagery

Author
item Yang, Chenghai
item Suh, Charles
item Westbrook, John

Submitted to: Journal of Applied Remote Sensing (JARS)
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/21/2016
Publication Date: 1/12/2017
Citation: Yang, C., Suh, C.P., Westbrook, J.K. 2017. Early identification of cotton fields using mosaicked aerial multispectral imagery. Journal of Applied Remote Sensing (JARS). 11(1):016008.

Interpretive Summary: Early identification of cotton fields is important for advancing boll weevil eradication progress and reducing the risk of reinfestation. Remote sensing has long been used for crop identification, but limited work has been reported on early identification of cotton fields. This study evaluated aerial imagery for identifying cotton fields before cotton plants start to bloom. Aerial color and near-infrared images taken over an 8 km by 12 km cropping area were mosaicked and then classified into different crops and cover types using image classification techniques. Results showed that classification maps were able to correctly identify over 90% of the cotton areas. The methodologies presented in this study will be useful for boll weevil eradication program managers to quickly and efficiently identify cotton fields at relatively early growth stages using mosaicked aerial imagery.

Technical Abstract: Early identification of cotton fields is important for advancing boll weevil eradication progress and reducing the risk of reinfestation. Remote sensing has long been used for crop identification, but limited work has been reported on early identification of cotton fields. The objective of this study was to evaluate aerial imagery for identifying cotton fields before cotton plants start to bloom. A two-camera imaging system was used to acquire red-green-blue (RGB) and near-infrared (NIR) images with 1-m pixel resolution along two flight lines over an 8 km by 12 km cropping area. The images were mosaicked using two approaches: manual georeferencing followed by position-based mosaicking in Erdas Imagine and content-based automatic mosaicking in Pix4DMapper. The mosaicked images were then classified into different crops and cover types using supervised classification techniques. Results showed that both types of mosaics were effective for cotton identification and that maximum likelihood classification produced the best overall accuracy of 90% for the position-based approach and 91% for the content-based approach. The methodologies presented in this study will be useful for boll weevil eradication program managers to quickly and efficiently identify cotton fields at relatively early growth stages using mosaicked aerial imagery.