Photogrammetry: A long history, from its invention to the most modern techniques
In 1849, a French army officer had the idea of using landscape photographs to observe and measure fields. If he can be considered as the inventor of photogrammetry, we must recognize that the development of the equipment like digital cameras, powerful computing units, intelligent software, and drones has drastically given new impulses to this area.
This technique, based on human perception of the terrain by stereoscopic observation, nowadays allows to make quite accurate measurements from aerial photos triangulated and georeferenced thanks to powerful photogrammetry software.
With the development of drones embedding high-resolution cameras, photogrammetry became an efficient and cost-effective method of data collection, whether the final application is stockpile volumes measurements, the generation of digital surface models (DSM), digital terrain models (DTM), digital elevation models (DEM) point clouds, contour lines, or orthomosaics.
Moreover, the miniaturization of sub-centimetric GNSS (Global Navigation Satellite System) receivers that can now be integrated into drones allow to eliminate the tedious task of placing and measuring ground control points (GCP) with quite heavy and costly topography equipment.
PPK (Post-Processed Kinematics) georeferencing consists of merging images with precise camera positions recorded at the times of triggering. These coordinates will be refined offline thanks to a specific software, either with RINEX (Receiver Independent Exchange Format) data downloaded from an external service, or from user’s own GNSS base receiver.
The photogrammetry software workflow will then process photographs containing accurate metadata, which reduces the manual intervention and so the overall processing time.
But cameras capture only the first visible layer. This is why in some operating conditions, users meet the limits inherent to photography, especially when there is vegetation, shadows or changing light conditions. This is where drone-based LiDAR (standing for Light Detection And Ranging) steps in.
Drone-based LiDAR for aerial surveying missions
Aerial LiDAR systems seems to be the answer to photogrammetry’s limits in some surveying and mapping applications.
If the early LiDARs that appeared on the market a few years ago where quite big and too heavy to be efficiently embedded on drones, the new emerging generation becomes much more adapted and in some cases even fully integrated to its unmanned flying platform.
Their know-how of the aircraft components like GNSS and IMUs, combined to their awareness of aeronautical constraints, brought the autonomous vehicles industry to develop particularly lightweight autonomous LiDARs.
They are actually in the best position to design the most optimized systems, integrating small IMUs and GNSS receivers easily combined to the sensors, and light enough to be carried by drones.
The first aerial LiDAR systems came out for multirotor drones, to which they are especially well suited. Actually, they can scan at a relatively slow rate with a limited range of about 50 m allowing to fly at low altitude over the area of interest, thus collecting a very high density of points.
For large areas, fixed wings and VTOL drones (standing for Vertical-Take-Off-and-Landing) allow to fly longer at faster air-speeds, thus enabling to survey wider areas in a single flight.
Today’s airborne LiDARs can typically scan at a speed from 15 to 100 km/h at 10 to 60 m high and collect up to 500 pt/m². Their weight is below 2 kg, and they are fully autonomous with their own IMUs, GNSS and RTK receivers.
LiDARs can collect accurate data whatever the light conditions are, and these data can be quickly post-processed thanks to their built-in PPK (Post-Processed Kinematics) functionality. Furthermore, LiDARs can capture points on the true ground even through vegetation. LiDARs also enable to represent complicated structures such as trees and power pylons in much more detail and with greater accuracy than 3D data reconstructed by photogrammetry technique.
All these advantages make aerial LiDARs ideal for many applications, such as corridor mapping, mining and quarrying management, forestry, civil engineering, environmental monitoring (like coastal erosion, for instance), archaeology, etc.
Will aerial LiDAR scanning replace photogrammetry?
Photogrammetry remains the most affordable method of data collection when used in the proper operating conditions (i.e. good light conditions and textured subjects).
The greatest advantage of photogrammetry is to deliver colored imagery, allowing to create realistic 3D models and orthomosaics comparable to satellite images but with much more details.
So one cannot predict that aerial LiDARs will totally replace photogrammetry, because both techniques are actually complementary.
What about the future?
If the data collected by LiDAR can be imported directly into any GIS software, the fact of producing uncolored images on the RGB scale can make it difficult to read for human eyes. This is why the trend becomes to combine point clouds collected from LiDAR together with photogrammetry in order to benefit from the advantages of both techniques, thus getting highly accurate data with RGB colors information.
This gives a complete representation of the environment, including both the dimensions and the textures, which is ideal for complex projects.