Detection of Vegetation Height using a Small Unmanned Aircraft System (sUAS) on Hawai’i

The recent proliferation of small unmanned aerial systems (sUAS) along with development of the photogrammetric processing techniques known as Structure from Motion (SfM) and Multi-View Stereo (MVS) have made the use of sUAS collected imagery more readily accessible to a variety of users. These sUAS platforms provide a mechanism for collecting overlapping imagery over targeted landscapes followed by production of dense 3D point clouds, digital surface models (DSMs), digital terrain models (DTMs), and high-resolution orthomosaics. For a study site located in the Pohakuloa Training Area (PTA) on the big island of Hawai’i, a vegetation height raster was generated by differencing the photogrammetrically-derived DSM and DTM surfaces. The photogrammetric aerial survey was conducted using a Skydio X2D quadcopter carrying a 12-megapixel RGB camera payload. Two imagery acquisitions were conducted at altitudes of approximately 60m and 120m above ground level (AGL) resulting in ground sample distances (GSDs) of approximately 2.3cm and 5.0cm, respectively. An accuracy assessment of the resulting vegetation height raster was conducted using in-situ measurements collected using terrestrial laser scanning (TLS) data. This paper presents an assessment of how well the technique of differencing photogrammetrically-derived DSM and DTM surfaces captures vegetation height, the effect of variable flying height on vegetation height determination, and the challenges of extracting vegetation height from sUAS-based photogrammetric surveys.