At 3Dronemapping, we often get asked how our data and end products compare to traditional LiDAR captured from manned aircraft. Each of the methods and instruments used to generate datasets have their own pros and cons. Certain project sites may be better suited to LiDAR while others would benefit from PhoDAR. This article hopes to indicate the differences between the methods of data capture and suggest where each is best utilized.
LiDAR (Light Detection and Ranging), is a remote sensing method that uses light in the form of a pulsed laser to measure ranges (variable distances) to terrain. This coupled with IMU and GNSS measurements allows us to reference the laser data and is adjusted to ground control points. Further data capture with a high resolution still camera allows for orthophoto generation. All of the above are transported in a manned aircraft, many of which are specifically designed for aerial survey and have the necessary rating from their respective aviation regulators. Such systems cost hundreds of thousands of dollars and are very susceptible to damage from dropping, dust, etc.
PhoDAR on the other hand is a slightly misleading acronym. PhoDAR is a portmanteau word that joins the words “photography” and “LiDAR.” The technology creates 3D point clouds by processing imagery. PhoDAR is another name for structure from motion (SfM) photogrammetry and is essentially how aerial surveys where done in years gone by using 9-inch film and very large metric cameras. No direct measurements are made to the ground, but this method uses GNSS and IMU observations like LiDAR that also directly geo-reference the data. This is the most common way of undertaking an aerial survey via drone
While both systems can be installed on either a manned or unmanned aircraft, airborne laser scanners are typically too heavy and power intensive as a payload for drone. The modern LiDAR sensors that are light enough for use on a drone are still too under-powered and suffer from limited range and poor return frequency meaning the aircraft must fly a lot lower and slower to make economic sense. These systems are also prohibitively expensive.
Ground control surveys are essential for both LiDAR and PhoDAR. Points for the referencing for the survey to the terrain and need to be evenly distributed over the site. Both systems use DG or direct geo-referencing with GNSS and IMU observations to perfectly accurate levels to initial create the survey block adjustment.
It is important to note that LiDAR is "better" than PhoDAR or visa versus. Each has their own strengths and weaknesses and each method are better suited to certain projects. Let us discuss the various pros and cons then see what system applies best to a few scenarios.
LiDAR Pros
Can cover large areas as the aircraft can efficiently cover more area than a drone
Multiple return pulses can "see" through vegetation
Data is collected in real time
Vertical accuracy of about 7cm
LiDAR Cons
Fairly expensive to survey, especially for smaller areas (<8 000ha)
Classification of terrain features is still done autonomously / manually
Pointclouds need to be coloured by imagery
Pointclouds are less dense compared to PhoDAR (1-3 points per m²)
Aircraft requires fuelling and can be a challenge in rural areas
Aircraft requires a tarmac runway nearby to reduce ferry
PhoDAR Pros
Very cost effective for areas up to 25 000ha
Quick deployment
Good for rural or difficult to reach projects
Can take-off / land anywhere
Generates an orthomosaic as a by-product
Very dense pointclouds (5-15 points per m²)
Vertical accuracy of about 7cm
PhoDAR Cons
Cannot cover large areas efficiently
Subject to strong winds
Cannot "see" into contiguous vegetation
Does not work well over homogeneous terrain like beaches
Classification of terrain features is still done autonomously / manually
Accuracy and end products are virtually the same for both methods. While not the subject of this article, orthomosaics from drones tend to be clearer and have greater colour depth since they are captured closer to the ground and have less dust or "air" to see though. Both data sets need to run though specialized pointcloud editing software to remove any outliers from high or low noise points in respect of laser data reflecting of dust / rain particles of from poor geometry fixes extracted from photogrammetry software. Both datasets then need to be run through various algorithms to further extract vegetation, buildings, bridges, water, etc.
LiDAR data capture is best utilized for large areas (no less than 8000ha) where time is a concern. It is good for collection of large datasets (mine exploration, flood risk analysis, national surveys) or a string of sites that can be capture in a single flight. It excels in densely forested areas where laser returns can penetrate gaps in the foliage. In areas where there is controlled airspace such as near airports, manned aircraft can much easily be directed by air traffic control.
PhoDAR does well for areas up to 25 000ha. More time will however be spent on the ground however to cover the same area. Remote sites such as islands, farms, mines are good places for PhoDAR if portion of the ground is visible between vegetation. Since drone surveys do not carry as many site establishment costs compared to manned aviation, project with smaller budgets can afford accurate surveys. As drones are portable, surveys can be done in very remote locations and very time efficiently. Since ground control surveys would need to be done in either case, only a single deployment is needed to do both ground control and the PhoDAR data capture.
It is important to not only chose the correct data capture method for your project but also to use a reputable company with experience and the correct equipment. Airborne laser systems that make up LiDAR and the aircraft are not cheap to own or operate. The companies that undertake these types of surveys are generally reliable and well established. However, there are many companies that offer inferior PhoDAR surveys by using improper aircraft, sensors and software. It is best to seek out a firm that has many years of experience in the field and has all the correct hardware with processing knowledge.