There are two main technologies that are used in aerial mapping with drones, photogrammetry and direct measurement with LIDAR. While both technologies can be used to create accurate maps, the path required to produce the final point cloud and surface vary. Considering that mapping projects have differing requirements, lets take a look at some of the benefits and differences of each technology to better understand their applications.
When deciding between mapping solutions, cost is often a factor. LIDAR mapping requires precise knowledge of the system's orientation to accurately perform direct measurement. The equipment required to accomplish this is what typically drives the cost of LIDAR above photogrammetry.
Size and Weight
LIDAR payloads are often larger and heavier than the imaging sensors required for point cloud generation with photogrammetry. This is largely do to the volume required to house all the required parts to take a direct measurement, including the rotating laser sensor, the inertial measurement system and required processing devices. LIDAR usually has to dissipate heat, which is why many solutions use metal cases with heat sinking structures. Camera technology on the other hand has had years to miniaturize and camera orientation is largely determined by processing software like Metashape or Pix4D. A sensor's size and weight is important because it determines the type of aerial platform that it can be installed on. On small drones, a large payload will decrease flight time and therefore mapping area.
Both LIDAR and photogrammetry solutions require some type of post processing, there is no silver bullet. Photogrammetry is the measurement of distances using images from different reference points which requires computing power. With the large images produced by today's cameras, upwards of 20 megapixels, the amount of processing can be time consuming. For large projects, special computers or cloud computers may be required. LIDAR processing however is much lighter and faster, since the measurements are taken directly in real time. LIDAR processing can consist of PPK positioning, point cloud classification, trajectory alignment or point cloud cleanup.
LIDAR is incredibly good at penetrating tree canopy, bushes and grasses. Most LIDAR sensors have multiple returns, which aids in finding the distance to actual ground, which is critical for some mapping applications. Photogrametric processing software, however has a hard time reconstructing tree canopy and it may be difficult to measure ground elevation when vegetation density is high. Another challenge for processing software is locating key points in dense canopy that it can use to reference between images. In mining applications, vegetation is typically of no concern, but in forestry, LIDAR may be the only option.
Since photogrammetry is based on aerial imagery, a light source is required to capture useful images. Therefore, data capture is limited to daylight ours. LIDAR however, produces its own light source, and can be used in any time of day and in most weather scenarios.
LIDAR sensors generally have a range limit based on their laser strength and reception sensitivity. This range can vary depending on the type and color of the surface that is being measured. Coal mining environments may reduce the range on a LIDAR payload, while photogrammetry does not suffer from range limitations.
Results vary depending on the LIDAR specifications but most LIDAR payloads will be able to quickly resolve small objects like cables, poles and fences. Photogrammetry on the other hand may require a lot of imagery from many angles in order to reconstruct thin objects. For this reason LIDAR is better suited for power line measurement and modeling.
As you can see each mapping technique has its strength and weaknesses. Contact us if you need to make an informed decision on what mapping tool to use.