Assessment of Turfgrass Health Using Aerial Images from Small Unmanned Aircraft.

Normalized Difference Vegetation Index (NDVI) and Dark Green Color Index (DGCI) are generally accepted as visual measures of turfgrass health but they are difficult to measure quickly over large areas of golf course greens and fairways. To this end, aerial images of creeping bentgrass turf plots were captured by small unmanned aircraft (UA) with a standard red-green-blue (RGB) camera and a modified near infrared (mNIR) camera over a 14 day study in which the plots were stressed by mowing at different heights. The mNIR camera was created by removing the internal NIR optical blocking filter from an RGB camera and adding a red pass optical filter to record red and NIR but block blue and green light. All images were captured at 50 feet above ground level (AGL) with 4000x3000 pixels to provide approximately 0.25 inch per pixel resolution. Five visual color indices - Normalized Green-Red Difference Index (NGRDI), Green Leaf Index (GLI), Visible Atmospherically Resistant Index (VARI), Green Chromatic Coordinate (GCC) and DGCI - as well as modified NDVI (mNDVI) measurements from the UA images for each of 27 plots were correlated to corresponding ground truth data collected using a handheld NDVI device. Interestingly, color indices produced better results (R²~0.8) compared to mNDVI correlations (R²~0.6). A novel method to calibrate RGB coefficients for color indices tailored to specific species is also presented.