Nischwitz, Claudia; Compton, Tyson (Utah State University)
Faculty Advisor: Nischwitz, Claudia (College of Science, Biology Department)
This research evaluates the use of Unmanned Aerial Vehicles (UAVs) in agricultural applications. We center our research on early disease detection and yield estimation in vegetable crops using aerial imagery and computer software. Previous research on UAV use in agriculture has addressed topics such as soil and field analysis (Long, 2017), Precision Viticulture in Italy (Matese, et al., 2015), and other areas pertinent to agriculturists. Our research builds on previous studies and aims to provide Utah farmers with knowledge and tools to increase agricultural productivity. A DJI Inspire drone is used with both a traditional light camera and a Near-Infrared (NIR) camera. Normal and NIR images are taken at the USU Research Farm in Kaysville Utah, and over local farm fields in Utah throughout the growing season. Unhealthy plants, identified from the aerial images, are tested at the USU Plant Pathology lab to identify diseases. Computer software (ImageJ, Microsoft ICE, and MATLAB) is used to process the images and collect crop health and yield estimate data. At the end of the growing season, the yield for each crop is measured and correlated to the aerial image data to create a predictive model for yield. Some plant diseases including Beet curly top virus in tomato and powdery mildew in squash are readily identified. We find that yield estimation with aerial imagery works well for specific crops. Potato yield was correlated with plant size at different numbers of days after planting. Further tests in coming years will provide validation for these results. Our current data show that the use of an UAV can be a valuable tool for early disease detection and yield estimation in vegetable crops.