Terrestrial Laser Scanners. Some Clouds for the Geomatic

In the second half of the 1990s, a number of specialist firms and laboratories introduced a number of unusual instruments that combined the principle of the surveyor's tacheometer with that of scanning a sheet of paper: a laser scanned the environment spherically to produce a three-dimensional point cloud, a 3D "raster" or "cloud".

The first results presented to the public were surveys of refineries, often complicated to map due to the intricacy of the piping. In its infancy, lasergrammetry (a technique dealing with measurements in the results of laser scanners) regularly offers its advantages for outdoor architecture.

Very soon, firms active in the field of topometry saw its potential for new activities (surveys of complex, "as-built" structures, etc.) and above all competition for total stations (motorized tacheometers with storage and calculation functionalities, the main tool of geomaticians at the time). They buy up some developments or create collaborations, for example :

• Leica Geosystems acquires Cyrax ;

• Trimble takes over Mensi ;

• Faro remains independent in its developments, which were initially more industry-oriented. Today, Trimble resells certain models under its own name;

• Zoller+Frölich has long enjoyed a close working relationship with Leica Geosystems.

Today, hardware developments have reached a certain maturity in terms of speed and precision. Current developments focus on filtering algorithms and automated point cloud assembly capabilities.

Make no mistake: obtaining a 3D cloud of an environment is just the beginning of the job. This raster can be processed in a number of ways:

• direct use of the 3D point cloud to visualize an environment, take measurements or integrate a project. This point cloud can be colored according to different information (laser reflectance, the color of a photo taken in parallel with the point cloud, etc.). The point cloud can also be used to maintain a snapshot of the current situation;

• after digitizing the cloud's characteristic points, we obtain 3D vectors, or even 3D shapes. This process, which is still very manual, greatly reduces the volume of data and highlights the essential elements that were the objective of the acquisition;

• modeling is the search for geometric shapes (best planes, cylinders, spheres...) or composite surfaces (mesh) in this cloud. This transformation into "average" vector elements limits the impact of measurement noise to obtain virtual models, the basis of many products (digital...