Land Surveys assisted Lachlan Shaw, in his Bachelor of Surveying Honours Project, in assessing the suitability of different surveying methods for monitoring the accretion (build-up of sand) by flying UAV mounted LiDAR and photogrammetric capture over the spit shown in Figure 1.
Figure 1. Land Surveys is assisting Curtin University to understand the unusual build-up of sand on the Peninsula, near Penguin Island.
LiDAR surveys use a pulse laser light emitted from a scanner. The reflected pulse is measured by the scanner and a distance is calculated for each emitted pulse based on the travelling time from emission to return. Whilst photogrammetric surveys rely on existing light conditions and edge/image matching to relate multiple images into a point cloud surface model, the ability of the LiDAR to directly measure the surface is what sets it apart. Features such as sand dunes and beaches can pose difficulties for photogrammetric surveys due to a lack of texture as a result of the homogenous nature of the sand surface, this results in a surface model which is noisier than normal making accurate measurements difficult. The photogrammetric survey was controlled using eight photo control targets evenly distributed every 50-100m throughout the site and surveyed using Real Time Kinematic GPS. Both the LiDAR and photogrammetric surveys were related back to the Australian Height Datum via a nearby State Survey Mark (SSM).
Figure 2. A typical LiDAR survey will output a colourised point cloud which can be used to generate contours or used as the basis for modelling and 3D feature extraction.
The LiDAR and photogrammetric surveys were compared to around 90 RTK GPS shots as shown in red in Figure 3 below. All three surveys were conducted within two to three hours of each other. The LiDAR survey taking approximately five minutes of flying and the photogrammetric survey about 15 minutes. The photogrammetric survey comparison indicated an average +8cm vertical error and LiDAR survey a +3cm vertical error over the 90 check shots.
Figure 3. Cross sections (in red) are the common methodology for surveying beach profiles, however, the results are less detailed and informative than the LiDAR survey.
A cross sectional comparison of the point clouds in Figure 4 highlights the photogrammetric model is high and significantly noisier than the LiDAR derived data. The photogrammetric survey also required eight more control points compared to zero for the LiDAR and yielded a result two times the vertical accuracy of the LiDAR data at four times the flying time.
Figure 4. Photogrammetric point cloud data (white) is typically ‘noisier’ than LiDAR data (brown).
In summary, this comparison shows that the LiDAR data captured using Land Surveys high end LiDAR system highlights advantages in both survey accuracy and reduced field times, making this methodology ideally suited to applications of monitoring accretion and erosion of coastal areas.