Skip to main content
ARS Home » Research » Publications at this Location » Publication #118516
Title: MULTISPECTRAL MULTITEMPORAL REMOTE SENSING FOR SPIDER MITE DETECTION IN COTTON

Author
item Fitzgerald, Glenn
item MAAS, STEPHAN - TEXAS TECH.
item Detar, William

Submitted to: International Conference on Precision Agriculture Abstracts & Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 7/18/2000
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Spider mites (Tetranychus spp.) are a serious pest in California cotton. Field scouting followed by pesticide application to an entire field is the usual method for detection and control of these pests. In this study, the Shafter Airborne Multispectral Remote Sensing System (SAMRSS) was flown over cotton fields to determine if spider mites could be detected early enough in the season for farmers to take corrective action. Images were acquired on 26 dates in 1998 and 29 dates during the 1999 growing season in the green (550 nm), red (660 nm), and near infrared (850 nm) wavelengths. Unprocessed images of these fields revealed little information about the status of mites in the fields. However, an image processing protocol was developed which included principal components analysis, pixel classification, and change detection. From the resulting image maps, regions of mite infestation could be identified shortly after they were detected on the ground. The same protocol was applied to data from both years. It was able to successfully identify mite infested areas within each field each year. Maps derived from remotely sensed imagery could be provided to field scouts to assist in locating spider mite infestations. This could facilitate precision pesticide applications to targeted areas of pest outbreaks. This has the potential to reduce the amount of pesticide applied which could lower costs to the farmer and benefit the environment by reducing the area extent of pesticide applications.