The implementation of a ground control point (GCP) network improves the accuracy of photogrammetric data. In order to get the most benefit of a GCP network please follow the suggestions below which are broken down into the following sections:

The number and the location of GCPs will greatly impact the usefulness of the network.

Location of GCPs

When planning your GCP network you should consider the perimeter, the interior and the highest and lowest elevation areas (if relevant).

GCPs should be spread uniformly around the perimeter of the site, whilst ensuring that at least they have at least a 50 metre buffer from the edge of the KML boundary (capture area). This will ensure a sufficient number of photos will contain the GCP allowing for maximum benefit.

A GCP should be placed in each corner of the KML boundary and then along the edges between the corners (see below about spacing). Irregular boundaries will require a more complex arrangement of GCPs around the perimeter.

Once adequate GCP coverage has been installed around the perimeter, GCP’s should then be spread uniformly throughout the interior of the site ensuring the high and low areas are captured (e.g. in a large pit it is advisable to have a GCP at or close to the bottom, as well as in the area surrounding the top). Please see the image below for an example of a suggested arrangement of GCPs at different locations and heights.

Density

It is recommended that a minimum of 7 GCPs be used and more added as a site gets larger. Please see the table below for a guide for the density and number of GCP’s required for a given site.

Note that this density accounts for a number of GCPs to be utilised as checkpoints to assess the accuracy of the generated model. Please refer to the Surface & Checkpoints section below for more information.

Ground control point targets may consist of physical markers/panels or targets painted directly on to the ground that can be accurately identified from the aerial imagery. In general it is recommended that each target be at least 50x50cm in size in order to make identification easy and accurate. Please see the image below for an example of a target that is somewhat harder to identify.

GCPs sprayed directly onto the ground should consist of lines at least 10cm in width, with a contrasting colour to the surface below.

GCPs should be placed on a flat surface free of any nearby obstructions in order to maximise their visibility. The centre of the GCP should be clearly identifiable by either the intersection of a chequered pattern or cross as this point is required to be identified in processing software to make use of the GCP network.

The images below show examples of suitable ground control targets:

A reusable and moveable chequerboard GCP A temporary spray painted GCP

A semi-permanent GCP spray painted on a concrete surface

Following the placement of a GCP the next step is to survey the point which is best done with a GNSS system consisting of a GNSS base and Rover. We advise using Continuous Operating Reference System (CORS), this means only needing a cellular device and one GNSS Rover; however without cellular coverage a GNSS base and permanent survey mark (PSM) will have to be utilised instead.

To validate the accuracy of the GPS measurements taken they should be compared against the measurement of a known position which may include:

There is also a free app that has access to all PSMs in Australia called Benchmrk which can be downloaded for iOS or Android here.

Along with the collection of GCPs, surface and checkpoint measurements are advised to be also captured throughout the site. Checkpoints should be marked on the ground using the GCP design suggestions above whereas level points require a point measurement of the ground but they do not require any marks on the ground.

These points are used to validate the vertical accuracy of the generated and can be captured en-route between GCP locations and are ideally located in the gaps between GCPs.