LAUSR

ABSTRACT Owing to the tiny, slim shape and clumping features of needles in a shoot, segmenting individual conifer needles or shoots using ground-based LiDAR (Light Detection and Ranging) is challenging. Very few measurements techniques or models have focused on the shoot point cloud. This letter presents a case study on detecting the dieback rate of individual shoots using LiDAR, which assessed the ability and sensitivity of LiDAR parameters to detect shoot shape and dieback rate. First, typical three-dimensional (3D) models of pine shoots were generated. Second, the waveform simulation model used in large footprint LiDAR was modified to simulate the small footprint LiDAR discretized returns on shoots. Third, a case study on Yunnan pine (Pinus yunnanensis) shoots was conducted to evaluate measurements using the Velodyne VLP-16 (Velodyne LiDAR Puck LITE-16) LiDAR. Finally, for the detection of shoot dieback, an expanded sensitivity analysis was performed on LiDAR beam divergence, angle resolution, and distance. The results suggest that it is nearly impossible to obtain 3D shape signatures of individual needles; however, shoot dieback rate can be extracted if mean intensity and point number features are properly used.