Inspections consist of a set of overlapping photographs of an entire roof or structure enabling the client to see / get a good view of a roof & it’s condition - at a particular point in time. Inspections provide full access to areas of a roof that are difficult or impossible to access by traditional methods. These images can include chimneys, valleys and guttering and are then available to the client, at any time, from the comfort of their office or desk.
Images are shot in high resolution with a general but lightweight camera drone delivering 48MB / 4K images and has the added advantage of digital zoom, for close inspection.


Example of Inspection images
A survey consists of a similar set of overlapping images but in a more accurate Photogrammetric form. These are far more precise in terms of positioning ie centimetre rather than just the metre accuracy of the general inspection images (above). Therefore these can be used for various measurements and in industry specific application like CAD (Computer Aided Design) or BIM (Building Information Modeling).
Photogrammetry is a technique that uses these multiple images to be able to produce a point cloud of data ie images and their precise location (cm accuracy) both in relation to each other, the site / location and it’s position on the earth’s surface.
These images require a specialist drone RTK (Real Time Kinematic Processing) or PPK (Post-Processing Kinematic), these are GPS correction / enhancing technologies that provides precise real time GPS location accurate data for processing in specialist survey-grade photogrammetry software.
Photogrammetry and geo-referenced data
Essentially, photogrammetry takes these overlapping images & location data and stitches them together to create a true digital 3D model of the physical world – a digital twin. This is achieved by cross referencing the highly accurate geo-referenced images from the drone and, where absolute (or as near as) accuracy is required, the signal from a base station, and a recognised standard earth co-ordinate system like WCS / UCS / EPSG etc.
Therefore, Georeferenced data is data that has been associated with a specific location on Earth). This process is called georeferencing, and it allows users to determine the precise location of any point on the data.
This photogrammetric process enables the production of various useful outputs
Point Cloud
The first part of this photogrammetry process is to generate a point cloud from the data (images & position). The point cloud can be processed in to a fully textured three-dimensional object (model).
The point cloud and its models can then be used to take measurements, calculate volumes, produce geo-referenced data for import in to CAD and other architectural and construction applications, generate 2D or 3D orthomosaics and 3D models of an object or surface. A 3D orthomosaic is essentially a digital twin of a real earth object.
Orthomosaics are images that are geometrically corrected for perspective, along with lens distortion and camera angle, to create a perfectly straight-down view of all objects in the frame as you would see on Google Earth, but at vastly higher definition / resolution).
Example of a survey grid (to collect the data (images & location)
Example of point cloud data
Example of an orthomosaic
Example of a 3D model (roof only)
Example of a set of inspection images from the same photogrammetry project
Example inspection image
Ends
Comments