MULTI-platform uas imaging for crop height estimation: Performance analysis over an experimental maize field

Unmanned aircraft systems (UAS) imagery-based crop height measurements are usually determined by mean of canopy height models (CHMs) without describing specific metrics in detail. In this paper, a crop height estimation method was proposed based on UAS imagery: 1) the centerline for each crop row was determined first on the CHM raster, 2) row polygons were then drawn according to the centerlines and predefined width. The length of each polygon depends on the actual row length, and 3) the percentile height was computed by using the signals in each delineated polygon from the CHM raster. In this method, the polygon width is adjustable in accordance with appearance and shape of different types of crops. The study was conducted over an experimental maize field using multiple UAS platforms and cameras at the Texas A&M AgriLife Research and Extension Center at Corpus Christi, TX, during the growing season in 2016. The results showed that different platforms can recover canopy height in a similar pattern using structure-from-motion photogrammetry while the height statistical results differed in scale at the maize field due to different camera resolution and flight altitude.