Drone Mapping 101: Photogrammetry vs. LiDAR

Posted on
3D Insider is ad supported and earns money from clicks, commissions from sales, and other ways.

Advances in drone technology have made them useful for several applications including advertising, filmmaking, aerial photography, industrial inspection, and surveillance. One field that has benefitted greatly from drone technology is mapping. By replacing manned aerial surveys with drone surveys, mapping services have become cheaper and less labor-intensive. The turnaround for drone mapping projects is much faster than the manned surveys done in the past. By automating certain aspects of the survey, drones also produce results are more accurate and repeatable.

In this article, we define what remote sensing using aerial surveys are and discuss the difference between the two technologies used for drone mapping: photogrammetry and LiDAR. Each technology has its own benefits and disadvantages, and the choice of which one to implement will depend largely on the type of mapping project being done.

What is remote sensing?

Drone mapping is a type of remote sensing, which is a catch-all term for any data collection method where there is not physical contact between the object being measured and the person or object doing the measurement. Remote sensing is a technology that is used in many fields that are mostly sub-disciplines of the Earth Sciences – including geography, hydrology, geology, ecology, and oceanography. It has also been used for community planning, land surveying, and military intelligence. Agricultural surveying using normalized difference vegetation index (NDVI) technology is another common form of remote sensing.

Data gathered from remote sensing projects can be processed, combined, and manipulated to generate a form of “geospatial” data – numerical data that is tied to certain locational coordinates.

Nowadays, remote sensing mostly refers to data gathering using satellite-based or aerial-based technology. The data gathered from these surveys can be collected from the surface and subsurface of the earth, the atmosphere, or even bottom of the ocean. More than taking aerial photos, satellites and aircrafts can be equipped with sensors that use different bands of the electromagnetic spectrum to measure infrared or ultraviolet energy. There are also sensors that can measure the ambient magnetic field or take point-wise gravity measurements.

Why use remote sensing?

Remote sensing makes it possible to gather data on locations that are otherwise inaccessible. It has been used to map and determine the topography of ocean floors, glacial features, and heavily forested areas.  Militaries around the world have used remote sensing to generate infrared maps and aerial images of dangerous and war-torn areas.

Remote sensing is also a more practical and cost-effective way to gather geospatial data on large tracts of land. Where traditional methods made use of costly and slow ground-based data collection, remote sensing jobs can be completed faster and will require less manpower. Most remote sensing technology offer the ability for partial or full automation, which makes the data gathering process more accurate and faster.

In summary, remote sensing technology makes it possible to gather data on places that are otherwise inaccessible or excessively dangerous for humans. By employing aircraft-based data collection, projects can be done much faster and cheaper than if it were done using ground-based data collection. The quality and accuracy of data can also be dramatically improved using the automated processes commonly employed by remote sensing methods.

What are ground control points?

A surveyor has the option of establishing ground control points (GCPs) to enhance the accuracy of maps created using remote sensing. These GCPs are pre-determined points where surveyors take accurate GPS measurements using handheld GPS receivers. This “ground truth” is used to calibrate the data collected via remote sensing. The number of GCPs and the density of the stations is dependent on the survey design. More GCPs will vastly enhance the accuracy and quality of maps generated from the survey but will take more time to complete and will be more expensive.

Active and passive remote sensing

It is best to start by making a distinction between passive and active remote sensing technologies to better understand the difference between photogrammetry and LiDAR.

Passive remote sensing refers to gathering data that is emitted or reflected by the objects being studied. Examples of this technology include measurement of radiometric data, infrared (or heat map) surveys, and aeromagnetic surveys. Photography is a form of passive remote sensing, as it is essentially a process of capturing the visible light reflected off various objects. Photogrammetry is essentially an expansion of drone photography, making it a form of passive remote sensing.

In active remote sensing, the aircraft or satellite emits a form of energy which is then reflected or scattered by the object being measured. The reflected energy is then measured by a separate sensor to gather data on the object’s shape, size, or location. RADAR systems use a form of active technology, as does mapping using LiDAR technology.

What is photogrammetry?

Photogrammetry is the use of a series of photographs to generate a map. This map is more than just a stitched together version of separate aerial images. By taking overlapping photos, photogrammetry can deduce the depth of individual features on the images. This is done by a triangulation method that is very similar to how the human eyes work – our eyes essentially capture two separate images of everything we see and stitch them together.

The equipment needed to do drone mapping via photogrammetry is relatively simple. Any drone with a high-quality camera can be programmed to take multiple overlapping images along a pre-defined flight path. Processing these images into a single coherent map is usually done by special photogrammetry software, such as Pix4D and PrecisionHawk.

The advantages and limitations of photogrammetry

The main benefit of doing a photogrammetry survey is the availability of cheap technology that can do the task. Even a mobile phone camera that is mounted on a drone can be used for photogrammetry given a few automated commands using special photogrammetry apps. With the relatively basic technology needed to do a photogrammetry survey, it has become the more commonly used method as it is easier to implement and is much cheaper.

Another advantage of photogrammetry is its ability to generate high-quality aerial maps using a full spectrum of colors. Since it works using visible light, the maps generated with photogrammetry using a high-quality camera look great. These 3D elevation models are a great marketing tool and are often used to create visually appealing advertisements, videos, and presentations.

In the past years, photogrammetry technology has experienced a sharp decline in use for remote sensing. Despite its ability to capture high-quality images, its reliance on visible light makes it limited to capturing images only as the camera can see them. This can be especially problematic when the goal of the aerial survey is to generate a topographic map of a heavily forested area. Instead of capturing the underlying terrain, photogrammetry will end up creating a map that shows only the forest canopy. Areas with limited visibility and a lot of shadows will also result in a low-quality model from photogrammetry.

What is LiDAR?

LiDAR stands for “Light Detection and Ranging”. During this type of survey, a LiDAR sensor mounted on an aircraft emits continuous pulses of non-visible light (with wavelengths between 1550nm and 1064 nm). These pulses of lights then bounce off the objects in the area being surveyed, after which they are measured by the LiDAR sensor. Since the speed of light remains constant, the time between light emission and reception can be effectively used to size and location of various objects in the survey area.

This is a highly accurate technology that is founded on the speed with which light pulses are emitted, and the accuracy afforded by the constant speed nature of light. LiDAR signals can even penetrate the ocean surface to gather data on the topography of the ocean floor. As you can imagine, LiDAR surveys generate a lot of data which then need to be processed by special mapping software.

A huge advantage of LiDAR is its versatility. Since it does not depend on visible light, a LiDAR survey can be conducted even at night, or even at areas with low visibility. LiDAR signals can also pierce through vegetation, allowing it to do mapping surveys in areas with thick forest canopy.

LiDAR is inherently a more accurate mapping technology than photogrammetry. The speed with which pulses of light are emitted and received by LiDAR sensors creates a very dense point cloud which the map processing software can analyze to identify individual objects. Most LiDAR sensors promise that they can generate models accurate to 1 meter. With the addition of high-quality IMU sensors and GNSS sensors that measure the location and rotation of the scanning platform, the accuracy of LiDAR models can be improved to the level of 1 cm. This accuracy can also be achieved by using establishing several GCPs. In terms of topography and land surveys, this is a level of accuracy that is incredibly valuable and hard to replicate.

In terms of data collection and processing, LiDAR surveys are also inherently simpler and faster. Photogrammetry surveys require a large overlap of about 60% to 90% to generate highly accurate models, which makes the data collection process longer. On the other hand, LiDAR surveys only require about 20% to 30% overlap to generate data that produces much more accurate models. In the data processing front, more powerful computers are needed to stitch together and smooth out composites of thousands of high-quality images. Processing of LiDAR data is mostly numerical in nature, making the data process less demanding and time-consuming.

The biggest limitation that works against LiDAR is its high cost. LiDAR sensors can cost between $50,000 to $300,000, and most industrial-scale applications prefer high-end models. The whole array of sensors needed for a LiDAR survey, including IMU and GNSS sensors, can be quite heavy. This means having to use a drone with a heavy payload capability, which is usually the higher price range. All in all, most LiDAR setups can cost up to $350,000 – far beyond what many casual drone pilots can afford.

It must also be noted that LiDAR sensors gather data without any color, resulting in a monochromatic model that is not so visually appealing and may be difficult to interpret. They are useful for creating Digital Terrain Models, and monochromatic models can be colored according to data such as elevation values. However, LiDAR models are not appropriate for applications that require visual assessment, such as inspection of property and surveying of crop health in a farm.

When is it best to use either photogrammetry or LiDAR?

For many applications, photogrammetry is the default option due to the much lower cost of the technology. High-quality camera drones are very affordable and accessible nowadays, as are the software needed to post-process the images. Photogrammetry is the more practical option for small-scale mapping projects. It is also especially effective when mapping open areas with good visibility such as bare mine sites or stockpiles.

If your desired output is a 3D model of the survey area, then nothing can beat the visual flair of a photogrammetry model. Photogrammetry can produce 3D models with a full range of color, further aided by the use of high-quality drone cameras. Photogrammetry aids applications that require visual assessment, such as inspection of road conditions or vegetation cover. If accuracy is a concern, the accuracy of a photogrammetry model can be greatly enhanced by increasing the number of ground control points.

On the other hand, if you are mapping areas with a lot of visual obstruction such as tall buildings or forest canopy, then LiDAR technology is the way to go. LiDAR is also the best option if your model needs to include very small objects such as pipelines or transmission lines. Elevation measurements are much more accurate with LiDAR, which may be valuable in generating Digital Elevation Models for road works and other construction projects.  LiDAR is also a much more versatile technology, allowing you to continue field surveys even at night or in poor weather conditions.

Is it possible to combine both technologies?

There is actually a third option: to combine both LiDAR and photogrammetry to generate a highly accurate full-color 3D model. This works by “draping” the composite photogrammetry images over the 3D model generated by LiDAR. Given the different degrees of accuracy of the technologies, doing this is not very straightforward and will probably require a GIS application to clip the spatial data together. Naturally, this is a very expensive process that is all but reserved for the most large-scale industrial applications.

Final thoughts

Drone mapping is one of the fastest rising applications of drone technology, aided by the maturation of the technology making it cheaper and more accessible. The use of drones for mapping represent a huge upgrade from the traditional methods of land surveying and using manned aircrafts for remote sensing. Drone surveys are faster, cheaper, and produce better quality data. With all these benefits, it is hardly a wonder why the technology has been so readily embraced by industries that require aerial mapping.

Photogrammetry and LiDAR are the two types of technology available for drone mapping, and each option has its own benefits and limitations. Photogrammetry is cheaper and produces full-color models but achieves lower accuracy and has more demanding post-processing needs. On the other hand, LiDAR is a much more expensive technology that produces highly accurate, albeit monochromatic, models. The choice of which technology to use largely depends on application – there are particular use cases that maximize the capabilities of each technology.