| An active area of research over the past several years has been the application of airborne light detection and ranging (lidar) in airport obstruction surveying. The primary objective in an airport obstruction survey is to accurately geolocate vertical objects, such as trees, towers, buildings, poles, and antennas, on and around airfields and in runway approaches. Previous studies have shown that airborne lidar is a potentially viable alternative to more expensive, time-consuming, traditional field and photogrammetric surveys for some airports and approach types. However, several problems remain unsolved. Due to lack of reliable, automated methods for extracting and attributing airport obstructions from lidar data, as well as stringent controls needed to ensure flight safety, extensive manual labor by skilled human analysts is typically needed. This, in turn, significantly reduces intended cost and time savings.;In this work, a new approach to extraction of vertical objects (airport obstructions, in particular) from airborne lidar data is developed and tested. The approach is specifically designed to exploit additional data provided by the latest small-footprint, full-waveform lidar systems, which digitally record the entire return signal from each transmitted pulse at high sampling frequencies. The primary steps in the approach include: (1) applying an advanced deconvolution algorithm to lidar waveforms, followed by georeferencing to produce very dense, detailed point clouds in which vertical structures of objects are well characterized; (2) voxelizing the lidar point clouds to generate high-resolution 3D grids (volumes) of lidar intensity values; (3) computing a 3D wavelet decomposition; and (4) performing vertical object detection and recognition in the wavelet domain.;The approach is tested using lidar waveform data collected with a commercial system in two Madison, Wisconsin, project areas. The reference data consist of field-surveyed vertical objects, including towers, antennas, trees, poles, and buildings, as well as high-resolution aerial imagery. The results illustrate the advantages of full-waveform data, volume representations, and multiresolution wavelet analysis for airport obstruction surveying and related vertical object detection applications. An additional benefit of this work is the demonstration of a highly effective and efficient workflow for lidar airport obstruction surveys. |
