LiDAR is the practice of using light or non-visible (e.g., infrared) electromagnetic radiation to detect and measure distance to objects. LiDAR stands for LIght Detection And Ranging.
You've probably heard of LiDAR before, but you may not really know much about it. My goal in this article is to explain on a conceptual level how LiDAR works.
Applications of LiDAR
LiDAR shows up in documentaries and news articles fairly often because it plays a role in so many scientific fields.
LiDAR is used in everything from agriculture to meteorology, biology to robotics, and law enforcement to solar photovoltaics deployment. You might see LiDAR referenced in reports about astronomy and spaceflight or you might hear about its use in mining operations.
LiDAR can image small things like historic relics for archaeology or skeletons for biology. On the other end of the spectrum, LiDAR can also image very large things like landscapes for agriculture and geology.
Aerial LiDAR image of Angkor Wat. Image courtesy of Ancient Explorers.
Then you have LiDAR systems that image moving objects or may be moving, themselves. These non-static systems may become the most common form of LiDAR as they are used for machine vision in military systems, survey aircraft, and prototype autonomous cars.
Yet other forms of LiDAR are not used to image solid surfaces at all: NASA uses LiDAR in atmospheric research while other LiDAR systems are designed to function and image surfaces underwater.
Clearly, this technology is quite flexible.
Similarities to SONAR and RADAR
So if LiDAR is LIght Detection And Ranging, is it similar to SONAR (SOund Navigation And Ranging) and RADAR (RAdio Detection And Ranging)? Yes, sort of. To understand how the three are alike, let us start by talking about how SONAR (echolocation) and RADAR work.
SONAR emits a powerful pulsed sound wave of known frequency. Then, by timing how long it takes the pulse to return, you can measure a distance. Doing this repeatedly can help you get a good feeling for your surroundings.
SONAR imaging used to locate a sunken vessel at sea. Image courtesy of the NOAA Office of Exploration and Research.
The key element of SONAR is the sound wave, but there are many different types of waves. If we use the same principal and change the type of wave from a sound wave to an electromagnetic wave (radio), you get RADAR (RAdio Detection And Ranging).
There are many different types of RADAR that use different frequencies of the EM spectrum. Light is just a special range of wavelengths detectable by the human eye, which is one of the things that differentiates LiDAR from RADAR. There are other differences between the two, but I'm going to focus on spatial resolution as the key difference.
RADAR uses a large wavefront and long wavelengths, giving poor resolution. By comparison, LiDAR uses lasers (tight wavefront) and much shorter wavelengths. The wavelength directly dictates the resolving power of an imaging system: Shorter wavelengths (corresponding to higher frequencies) increase the resolving power.
Weather RADAR imaging picking up the movement of weather systems and migrating birds. Image courtesy of NASA.
And if your wavefront is huge, you have to wonder—what distance are you measuring?
Imagine trying to use a flashlight range finder for your next home improvement project. Is that flashlight measuring the distance to that beam or the wall behind it? This is too imprecise to be practical. Laser range finders, however, are much more accurate.
Measuring with Pixels
LiDAR uses a laser to measure distance, but that is not any different from a laser range finder—you point it at something and you measure a distance. But what if that one distance measurement were interpreted as a single pixel? You could then take numerous distance measurements and lay them out in a grid; the result would be an image that conveys depth, analogous to a black-and-white photograph in which the pixels convey light intensity. Sounds pretty cool, and that could be pretty useful right?
But we still have several questions to answer before we can fully understand how this system works.
How many pixels do we need?
If we compare LiDAR to early digital cameras, we would need one or two megapixels (or one to two million pixels). So let's say we need two million lasers, then we would need to measure the distance indicated by every one of those lasers, so another two million sensors, and then circuits to do the calculations.
Maybe a whole lot of lasers at once is not the best method. What if, instead of trying to take a whole "picture" at once, we did what scanners do? Namely, could we take part of the picture and then move the pixel and take the picture again? This sounds like a much simpler device that could be implemented with only one laser and one detector. However, it also means that we cannot take an instant "picture" the way we can with a camera.
How do we move our pixel?
In a scanner, the physical pixel is moved along the image. But that would not be practical in many situations, so we probably need a different method.
With LiDAR, we are trying to take a "picture in depth", or you could say we are trying to take a "3D model".
If you were paying attention in your math classes, you may recall the Cartesian coordinate system (X, Y, Z) and the spherical coordinate system (r, theta, phi). I bring this up because, if we assume that the LiDAR is at the origin in a spherical system, then we just need to know the horizontal angle (theta), the vertical angle (phi), and the distance (r) to build our 3D model.
Representation of a spherical coordinate system. Image courtesy of Wolfram MathWorld.
With only one source, the 3D model would be a single surface. But if we combine multiple surfaces mapped together, you can get a full 3D model. So we need one laser and one sensor to make our image using the spherical coordinate system.
Some LiDAR systems are mobile and use GPS or other positioning systems to map all the readings together into a single image. These mobile systems still use the same principles but they may be applied in different ways.
How do we actually change our laser's angles quickly?
We have two million "pixels" to measure. How can we possibly adjust our laser to measure them all?
There are only 86,400 seconds in a day, so working with any length of time over thousandths of a second to measure and adjust is probably too long to be practical.
So how do they adjust the pixel so fast? The key is that they never stop adjusting it, using a spinning mirror or prism rotating at a very precise and well-known speed. The laser is reflected off of that mirror or prism so that its position is constantly changing but at a known rate. This makes it very fast and easy to adjust one of our angles, either phi or theta, to scan across our image. For scanning the other angle, we could use a much slower system such as a precision stepper motor.
But we still have one last major problem to solve.
How do we measure the distance?
There are multiple methods of measuring distance with a laser, depending on what the system is trying to achieve. And a single system may be using multiple methods simultaneously to increase accuracy. All methods require very precise equipment.
The simplest to understand is time of flight (ToF). This is also most often listed as the method used to measure distance with a laser. If you calculate the time it takes for light to travel 2mm, the distance needed for 1mm resolution, you get the time 6.67 picoseconds. That requires some very precise equipment to measure, but can be done for a price.
Another method, triangulation, uses a second spinning mirror to redirect the receiver signal; by measuring the change in the angle, you can get a distance. Triangulation may not be applicable for long distances, and the use of the rotating mirror may complicate the system.
Finally, by modulating the laser, it is possible to measure a phase shift in the laser modulation. Because of the periodic nature of the modulation it cannot by itself be used to measure the distance. Rather, it produces a list of possible distances that can be used in addition to another method to increase the accuracy.
LiDAR rendering of a group of people. Image courtesy of NASA.
The LiDAR Laser
One last thing to talk about: the laser itself. Despite the name, most LiDAR systems employ infrared rather than visible radiation.
Because electromagnetic radiation's interaction with matter is governed by wavelength, some wavelengths work better for certain applications. Someday microwave (maser) or x-ray (xaser) lasers could be used to build LiDAR systems that can image through a wide variety of materials, greatly enhancing their usefulness.
With that, you have a conceptual understanding of how LiDAR works.
Obviously, LiDAR has not yet reached its full potential and I'm personally excited to see where this technology will go in my lifetime.
Special thanks to Aeromapss for the thumbnail picture.