POTENTIAL FOR REMOTE SENSING FROM AGRICULTURAL AIRCRAFT USING DIGITAL VIDEO

A field imaging system using low-cost digital video was evaluated for use on agricultural aircraft. The system uses a digital video (DV) camera, Global Positioning System (GPS), and Video Mapping System (VMS) for image georeferencing. Remote control and monitoring was implemented to allow the pilot to image specific field areas of interest, facilitating image acquisition and post-processing. Two practical applications are described to exemplify the use of digital video in agricultural aircraft. Broadleaf weeds were distinguished from cotton, and chlorophyll a (chl a) concentration was regressed against digital numbers obtained from images of catfish ponds. For the weed detection study, a mixture of spotted spurge (Euphorbia maculata L.) and hyssop spurge (Euphorbia hyssopifolia L.) were distinguishable from both early cotton (Gossypium hirsutum L.) and Johnsongrass (Sorghum halepense (L.) Pers.) using supervised classification algorithms. Results comparing a Mahalanobis Distance-classified image with the raw RGB image indicated cotton, spurge, and Johnsongrass correctly classified with 73%, 67%, and 82% accuracy, respectively. For the pond study, statistically significant relationships were observed between digital numbers from pond images and chlorophyll a (chl a ) from water samples obtained at multiple ponds. These experiments demonstrate the potential benefits of using low-cost digital video on aircraft to assist agricultural producers in improving management decisions.