Utilizing UAS-Lidar for High Throughput Phenotyping of Energy Cane

Uncrewed Aerial Systems (UAS) equipped with digital cameras and sensors is an effective remote sensing tool for High Throughput Phenotyping (HTP) in precision agriculture. While studies have established relations to estimate crop height and biomass from UAS data, there has been limited work that examines the relationship between field measured biomass to UAS-Light Detection and Ranging (lidar) estimated biomass for the energy cane crop. This study explored the utility of UAS-lidar for phenotyping energy cane crops. The study collected lidar and ground truth data from an energy cane experimental plot in Weslaco, Texas-USA. Random Forest (RF) regression analysis showed high correlation between modelled crop height from lidar and field measured crop height (r
²
= 0.94, rmse = 0.12 m, me = −0.002, mae = 0.009, n = 400). Also a RF model between field measured biomass and modelled crop height, point cloud density, intensity, and number of returns generated from the lidar showed high performance (r
²
= 0.85, rmse = 92.00 g/m
²
, mae = 76.11, me = 0.50, n = 300). These results buttress the capability of UAS-lidar for high throughput phenotyping as has been reported in other studies.