Mapping & Modelling Case Study

The client gave an open brief to map and model the 2.8km course with a view to future use as a mapping / visualisation tool, and to assess the potential for other business areas such as site survey, planning applications etc.

Due to the significant change in elevation over the length of the course, the area was split into 6 flights, as illustrated below.

We will detail the Base Station (flight 6) only, in order to provide a more focused case study.

A full pdf copy of the case study can be downloaded here.

Workflow

• Flight Planning
  • Search map base on Flight Mapping Software
  • Define area of interest to be mapped using boundary box 
  • Set inputs for altitude (AGL), image overlaps, flight path heading etc
• Flight
  • Perform preflight checks, then proceed with autonomous flight path
  • Monitor flight progress (change battery if required and resume flight)
  • Autonomous landing or switch to manual on return to home
• Generation of Orthomosaics & 3D Point Cloud
  • Upload geo-tagged images from on-board SD card
  • Define processing parameters
  • Process images, orthomosaics and point cloud using software (desktop or Cloud)
• Analysis / Production of Deliverables
  • View / edit orthomosaics, DSM, Point Clouds, etc
  • Analysis and measurement of volumes, stock piles etc. as required
  • Export of output files as required by client (geoTIFF, obj, dxf, LAS, KML etc)

Flight Planning

The flight planning software has a box bounded by points that can be adjusted to fit the area of interest, as shown below. The flight parameters, such as altitude and image overlap, are then specified and the software calculates the flight path, time and the number of batteries required.

Site Preflight Checks

• Weather checks, site survey and risk assessment 
• Physical checks of UAV system
• Survey take off / landing site for foreign object debris (FOD)
• Power on and calibrate UAV
• Ensure image transmission is being received
• Final weather check
• Conduct 360 check to ensure area is clear of others
• Take off to height of approximately 3 meters
• Check that UAV responds to commands
• Commence autonomous mapping flight

Monitoring the Flight

• Pilot will monitor the flight constantly for any changes in conditions, signal strength, battery power, persons entering site etc.
• Pilot should also confirm and monitor data / image  collection.
• Should a battery change be required, the UAV returns to home automatically, the battery is changed, and data collection will resume at the point where it left off.
• Following completion of the mission, the UAV will return to home automatically, but it is preferred to switch to manual for the landing.

Generation of Mosaics & 3D Point Cloud

During the Initial Processing the downloaded images are used to complete the following tasks:

• Keypoints extraction: Identify specific image features as keypoints.
• Keypoints matching: Find which images have the same keypoints
• Calibrate the internal and external parameters of the camera.
• Locate the model by geolocation information (images are geotagged).
• Automatic Tie Points are created during this step. 

Point Cloud and Mesh processing will build on the Tie Points with:

• Point Densification: Additional Tie Points are created based on the Automatic Tie Points that results in a Densified Point Cloud.
• 3D Textured Mesh: Based on the Densified Point Cloud a 3D Textured Mesh can be created.

DSM, Orthomosiac and Index processing:

• Produces the required outputs as defined by user such as DSM, 2D orthomosaics and 3D mesh models 
• Generates the project folder structure and files

Once the project is complete you can then look at the analysis and production of deliverables.

Analysis / Production of Deliverables

Once the processing is complete, you can then produce, edit and optimise digital models, point clouds, orthomosaics and measure distances, areas and volumes.

rayCloud:

The rayCloud (shown below) is presented in 3D and can be used to visualise the different elements of the reconstruction (Camera Positions, Rays, Tie Points, Densified Point Cloud, 3D Textured Mesh, Video Animation Trajectories) and their properties by checking / unchecking the box.

Automatic Tie Points:

An Automatic Tie Point is a 3D point, and its corresponding 2D keypoints, that were automatically detected in the images during initial processing and used to compute its 3D position. Each point will appear on a number of the images (as shown below), with each individual image producing thousands of tie points.

The right sidebar, on the Images section, shows all the images in which the point appears, with the tie point below being visible in 26 images.

Points Clouds :

Checking the Point Cloud box “inserts” the additional ties points that were created during the processing phase (based on the Automatic Tie Points) and results in a Densified Point Cloud. These images could be mistaken for video or photographic images, but are in fact point clouds.

Measurement

The software can be used to accurately measure distance, area and volume. The images in the right sidebar can be zoomed in to pixel level and used to adjust the position of the measurement points. This can be done over a number of the images to increase accuracy further.

Distance:

In the example of the Skoda Yeti in the carpark, the length in meters is measured as 4.22m and the dimension in the data sheet is 4222mm.

Areas:

Areas can be accurately calculated in a matter of minutes. The new area icon is clicked and points around the area are generated with mouse clicks. The images in the right sidebar can be zoomed in to pixel level and used to adjust the position of the measurement points. This can be done over a number of the images to produce mm accuracy.

Volumes:

As with distances and areas, volumes can be accurately calculated in a matter of clicks. The area is first defined from a top down view and the volume is then computed by the software. The settings allow for a variety of base surface option dependant on requirement. In the example below “the lowest point” was chosen to estimate the volume of material required to be removed to clear the area to ground level.

Mosaic Editor

The Mosaic Editor is used to visualise and export the orthomosaic map (below) and the Digital Surface Model (below). It can also be used to improve the visual aspect of the orthomosaics.

Output Files, Software & Uses

Outputs can be exported in the following formats

• Densified Point Cloud - las, laz, xyz & ply
• Grid DSM - las, laz & xyz
• 3D Textured Mesh - ply, obj, fbx & pdf

Some of the main uses are:

• 2D & 3D digitisation & visualisation
• Digital measurement of areas and volumes
• 3D printing
• Web viewing

Typical software use for GIS, CAD & Web Applications:

• Argis
• Gobal Mapper
• AutoCAD
• Rhino
• Google Earth
• Sketchfab

For a full list of outputs, uses and software please click the following link - Outputs, Uses & Software

A full pdf copy of the case study can be downloaded here.