Lidar receivers with exquisitely sensitive Geiger-mode detectors are able to detect surfaces even when the line of sight from the lidar sensor to the surface is highly occluded by intervening forest canopy. Additionally, repeated scanning of a region of interest from a diversity of perspectives increases the likelihood of imaging any given surface through at least one substantially unoccluded line of sight. Together, these techniques allow airborne lidar collections to be tailored to achieve comprehensive human activity layer (HAL) data collection, even in areas with dense foliage. We present a study of the performance of a 3DEO lidar for foliage poke-through applications, exploiting both its Geiger-mode sensitivity and agile geo-referenced scanning system. We present two methods for estimating the utility of the resulting 3D point clouds in the HAL, near the ground, based on the spatial statistics of the point clouds. We apply those methods to airborne Geiger-mode lidar data of deciduous forests in Massachusetts and conifers in the US Pacific Northwest. We quantify the completeness of the point clouds as a function of the collection parameters. We then use this analysis to estimate the ideal collection parameters for a Geiger-mode lidar with georeferenced scanning to yield a high-utility data product.
|