Aerial Surveying Deliverables

Providing you with the right deliverables is key to a successful aerial surveying mission. Most of our clients use a variety of software from AutoCAD Civil, ArcGIS, 12D, specialised spectral software or sometimes the deliverables we provide are the final data in a report generated in Microsoft Word. What ever the end requirements, we ensure that the files are compatible with. In addition to the file formats, there are also considerations of the type of post processing the NQ UAV is to be engaged to conduct. Some of the more popular outputs include a DSM (Digital Surface Model) or a DTM (Digital Terrain Model). A DSM is a model that comes almost straight out of the initial processing. It will be “lumpy” and contain objects such as: trees, cars, power poles, power lines, buildings, and most above ground synthetic structures (this includes bridges etc.). A DTM has either removed, or attempted to remove all objects that are not “bare earth”. The image below shows a quick comparison between the two.

Surfaces represented by a Digital Surface Model include buildings and other objects. Digital Terrain Models represent the bare ground.

The DTM can also be produced in 3 distinct ways:

  • Fully autonomous classification;
  • Semi autonomous classification;
  • Manual classification;

Depending on the requirements of your deliverables, the surrounding area and how quickly you need the data delivered, it will change which option is applicable for your project. In most cases Fully Autonomous will do exactly what you need, however in past projects, we have seen it smooth out sharp edges of synthetic waterways (concrete storm water drains), walls of quarries and open pit mines and some natural cliffs and bluffs. If these artefacts are not desirable in the aerial surveying deliverable, we move to semi autonomous classification. The classification process produces a point cloud where sections of the cloud are assigned to groups. These groups can be the filtered further or removed from the deliverables.

Semi autonomous classification allows us to input some features (such as walls of quarries and open pit mines) this typically produces a more accurate DTM and the features can be reliably identified prior to the operation commencing. It is also possible for the client to provide the features prior to the operation, please discuss this with us prior to ensure that information provided is compatible with our systems. Semi autonomous works best in smaller discrete areas where it is highly important that features are kept during the next stages of the DTM creation.

Manual classification is the most labour intense, it is also the most accurate, allowing our skilled GIS and data processors to identify each area of the point cloud and assign it to the corresponding classification. This is only used for difficult datasets that are in locations such as densely populated areas that have above ground power poles and power lines, dense buildings, and the terrain below changes elevation throughout the area of interest.

Orthomosaics are the other common deliverable, our Orthomosaics or Orthophotos can be delivered at full resolution (typically less than 3cm per pixel) or can be decimated to allow for easier file handling and storage. Depending on the platform that was used to capture the data, there may be as many as 250 “layers” from a hyperspectral camera or sensor.

Other treatments and processing considerations that NQ UAV routinely do during the processing stages include:

  • Gridding and decimating;
  • Colour balancing;
  • Contour Lines Generation;
  • Reflectance maps;
  • Tree counts;
  • Vector files;

Contact us today to discuss your deliverable requirements for your surveying project with NQ UAV.