What is a point cloud? 

Point clouds are data files created by LiDAR scanners. They are made up of millions, or even billions, of “points” which are essentially reflections off a surface picked up by the scanner. 

A LiDAR scanner works by firing pulses of light (LiDAR stands for Light Detection and Ranging), and measuring how long those pulses take to bounce back after hitting an object. Using the speed of light and that time measurement, the LiDAR works out how far away from the scanner each point is. 

By sending out millions of pulses, a LiDAR scanner quickly builds up a detailed picture of the environment around it. Some LiDAR scanners only provide 2D scans, which look more like floor plans. Others produce a 3D point cloud. Additionally, some scanners are static – they only fire pulses in front of them – while others spin, sending pulses all around the scanner to build a 360° point cloud. 

There are three things you can do with a point cloud: 

1. You can use it to measure distances between various points and your scanner (and therefore whatever it is mounted to). 
2. You can use it to measure distances between two points within the cloud. 
3. You can use it to create a model of the environment around you. 

Because of the volume of pulses sent out by a LiDAR scanner, the measurements you can take from a point cloud are generally very precise and very reliable. From those core uses for point clouds, you can do a number of things with the data. 

By mounting a LiDAR on a rover or an aerial vehicle, you can survey large areas of land quickly and in great detail. Surveyors can then use that data to take accurate measurements for a range of uses. A popular use of LiDAR, for instance, is to survey a space that an autonomous mobile robot will operate in. The point cloud can then be transformed into a map that can be fed into the AMR’s control stack, so it can plan a route through a space and navigate there without crashing. 

To use point clouds for surveying and mapping, it’s vital that the point clouds are georeferenced – we’ll come on to that later. 

By mounting a LiDAR on a rover or an aerial vehicle, you can survey large areas of land quickly and in great detail. Surveyors can then use that data to take accurate measurements for a range of uses. A popular use of LiDAR, for instance, is to survey a space that an autonomous mobile robot will operate in. The point cloud can then be transformed into a map that can be fed into the AMR’s control stack, so it can plan a route through a space and navigate there without crashing. 

To use point clouds for surveying and mapping, it’s vital that the point clouds are georeferenced – we’ll come on to that later. 

As we’ve said, surveying and mapping applications need a georeferenced point cloud to work. That means that each point must have a defined position, either in relation to the earth (in a global frame) or in relation to a known fixed point (a local frame). This allows users to put the point cloud data in context, so it can be used. 

To georeference point cloud data, you need a device attached to the payload or platform the LiDAR is mounted on that provides position data. That’s where hardware like our GNSS/INS comes in. 

A GNSS/INS is a device that combines data from GNSS satellites and an inertial measurement unit to provide an estimate of the GNSS/INS’ position (and the position of whatever it is attached to) on the earth. A GNSS/INS is designed to give you accurate and robust position data – which you can combine with your LiDAR data to georeference your point cloud. 

OXTS has a range of different GNSS/INS devices that you can use for georeferencing point cloud data. Some are designed for aerial mapping activities, such as our xNAV650, while others are designed for pinpoint accuracy including our RT3000 v4.  

See our full range of GNSS/INS devices 

At OXTS we’ve developed tools that help make it easier to georeference point clouds, as well as some advanced tools that let you get even more value from point cloud data. 

OXTS Georeferencer

OXTS Georeferencer is built specifically to combine location data from your GNSS/INS with raw LiDAR data, to produce a georeferenced point cloud. You load up your LiDAR data and your GNSS/INS data, press go, and the system takes care of the rest. The final point cloud output can then be used in a range of mapping software and tools. 

Boresight Calibration Tool

One of the biggest challenges when georeferencing point cloud data is eliminating blurriness and “double vision” from the point cloud (objects appearing twice). These issues are caused by slight misalignments between the LiDAR and the GNSS/INS, and the OXTS Boresight Calibration Tool fixes them. It requires around two minutes of extra driving at the start of a survey – but it dramatically improves the quality of your final point cloud.

LiDAR Boost is a technology designed to improve accuracy in GNSS-denied spaces. It works by analysing individual snapshots from an automotive LiDAR, identifying planes that indicate things like walls and corners, and comparing how they move between those snapshots. By working out how far they move and how they move between two snapshots, LiDAR Boost can calculate both forward and angular velocities, which can be fed into the GNSS/INS as an additional data source. 

LiDAR Boost is great for mapping activities taking place in GNSS-denied spaces like underground car parks. Compare the route mapped out by an INS without LiDAR Boost, to the same route mapped using the technology: 

We hope this blog has been a good introduction to point clouds and why they matter for surveyors and autonomous navigation engineers. The key takeaway is that point cloud data needs accurate georeferencing in order to deliver the most value – and that requires an accurate positioning solution. OXTS has a combination of hardware and software tools that enable mobile mappers to get accurate position data, even in environments where GNSS signal is weak or nonexistent; working with us, you can maximise the value of your point cloud data. 

If you’d like to discuss how OXTS can help you integrate point clouds into your workflow, click below to get in touch.