Drop #355 (2023-10-18): LiDAR, LiDAR, 👖^H🌳 On 🔥

LiDAR Primer; Visualizing LiDAR Data; CATS On Plumes, Wildfires, And Hurricanes; 🌕 LiDAR

Yesterday’s 🏥 bombing, and fog-of-war blame tossing for it has your friendly neighborhood hrbrmstr in less than stellar spirits, today. I suspect that may be the case for many readers, so I put together some more “exploring” resources to help you get lost in less soul-crushing thoughts.

While there are (thankfully) no point clouds of actual 👖 burning, one of the resources does show how LiDAR is used to get a unique perspective on wildfires.

Apart from a quick introductory section on LiDAR, the remaining three sections will follow the idiom of the previous “exploring” post and lightly introduce you to the resource, so you can spend more time exploring.

As one might expect, there's an EPIC R package for working with LiDAR data.

TL;DR

This is an AI-generated summary of today's Drop.

Perplexity must have gone out drinking, last night, as it failed miserably — TWICE — on the same prompt I’ve been using for many weeks. I ended up spending ~two minutes cleaning up the results. Trusting these giant systems for reproducible workflows seems pretty daft. I’m glad more finely tuned LLMs and GPTs do a much better job in focused contexts.

1. LiDAR Primer: This section is an introduction to LiDAR (Light Detection and Ranging), a remote sensing technology that uses laser light to measure distances and create detailed, three-dimensional maps of the environment. The author explains the basic principle of LiDAR and its applications in various industries, including automotive, infrastructure, robotics, and mapping.

2. Visualizing LiDAR Data: This section focuses on visualizing LiDAR point cloud data, pointing to a solid piece that shows how to visualize LiDAR data from the publicly available KITTI dataset.

3. CATS On Plumes, Wildfires, And Hurricanes: This section discusses the use of NASA's Cloud-Aerosol Transport System (CATS) in studying volcanic eruptions, wildfires, and hurricanes. The author highlights how CATS data helps model volcanic plumes and monitor wildfire smoke affecting air quality in the US. The primary resource for this section is NASA's Scientific Visualization Studio.

4. Moon LiDAR: This final section quickly directs readers to Moon LiDAR, so they can explore the 6 billion point NASA LOLA dataset.

---

LiDAR Primer

Photo by Scott Blake on Unsplash

While I could just leave you to read the Wikipedia link in the preamble, you deserve better from me than that. If you are already familiar with LiDAR, I'm likely not going to tell you anything novel, here, so feel free to jump to the resources.

LiDAR is an acronym for “Light Detection and Ranging”. It's a remote sensing technology that uses laser light to measure distances and create detailed, three-dimensional maps of the environment. It works by emitting laser pulses and measuring the time it takes for the reflected light to return to the sensor. This time is then used to calculate the distance the light traveled, and repeating this process millions of times per second creates a real-time 3D map of the environment.

The basic principle of LiDAR is pretty easy to grok: you take a laser light, point it at a surface, and measure the time it takes for the light to return to the source. This time, known as the “time of flight”, is used to calculate the distance to the object. The formula for this calculation is: distance = (speed-of-light x time-of-flight) / 2.

A typical LiDAR system consists of four main parts:

• LiDAR sensors

• GPS receivers

• a scanner and optics, and

• a computer system

The LiDAR sensors emit laser pulses, commonly in green or near-infrared bands, towards the ground as the airplane or helicopter travels. On vehicles (or humans), the pulses are emitted outward. The GPS receivers are used to track the altitude and location. This is especially important for accurate terrain and elevation values (in an aerial context). The scanner and optics direct the pulses towards the ground, and the computer system processes the data.

LiDAR systems can emit over 160,000 pulses per second, creating millions of data points that form a “point cloud”. This point cloud is used to generate precise 3D models. Depending on the need, different points can be filtered out. For example, in a forested area, LiDAR can reflect off different parts of the forest until the pulse finally hits the ground. This allows for the creation of images that depict only changes in ground elevation, which can be particularly useful in fields like archaeology.

LiDAR technology is used in many industries, including automotive, infrastructure, robotics, trucking, UAV/drones, industrial, mapping, and many more. It is particularly important for autonomous vehicles, as these vehicles need to quickly develop an image of the world around them to avoid hitting pedestrians, animals, obstacles, and other vehicles.

Compared to other similar technologies like radar (which uses radio waves) and sonar (which uses sound waves), LiDAR is generally more accurate because it uses laser light, which has a very short wavelength and is therefore able to provide more precise measurements. However, LiDAR systems typically have a shorter range than radar systems and require a clear line of sight, while radar can detect objects through fog, rain, and other obstacles.

GISGeography has a more visual 'splainer for those that process information better in that form.

OK. Enough of me blathering. Time for the good stuff!

Visualizing LiDAR Data

LiDAR sensors have become very affordable, with capable models available for well under $10K USD. As a result, their use has skyrocketed, and all sorts of crunchy LiDAR data is available to play with.

Alex Staravoitau visualizes some of this LiDAR point cloud data from the publicly available KITTI dataset, collected from a vehicle equipped with LiDAR and cameras.

Dependencies like NumPy, Matlab, and PyKitti are used to work with the point cloud data in a handy notebook and parse labeled objects. Sample camera frames show road features like tram tracks and parked cars.

Each LiDAR frame visualizes the laser beams and captures car and tram silhouettes marked with bounding boxes. Projections of the point cloud onto different planes reveal useful views, like a bird's-eye perspective. Animating the 114 frames over 11 seconds shows how the point cloud changes over time and makes labeled objects like trees and parked cars easier to interpret. LiDAR plays an important role in applications like simultaneous localization and mapping for vehicles.

Alex's post is very accessible, and the animation is somewhat mesmerizing. It's a solid, more technical/practical introduction to this tech, which should help folks appreciate the last two sections even more than y'all likely would.

CATS On Plumes, Wildfires, And Hurricanes

Photo by Amy Baugess on Unsplash

NOTE: The first NASA link intermittently dies.

NASA's Cloud-Aerosol Transport System (CATS) instrument on the International Space Station measures aerosols and clouds to study volcanic eruptions, wildfires, and hurricanes. CATS data helps model volcanic plumes and their long-range transport, which is super important for aviation safety. It also monitors wildfire smoke affecting air quality in the US.

In 2015, CATS observed the Calbuco volcano eruption (that's not the cool link) in Chile and wildfires in Oregon. It further tracked Hurricane Matthew and detected dust plumes from Africa traveling over 4 km high.

CATS uses multiple laser wavelengths to distinguish between aerosol types like dust and smoke. This demonstrates CATS' ability to rapidly characterize atmospheric events from space within 6 hours of observation.

There's tons more info and some super cool animated visuals over at NASA's Scientific Visualization Studio.

🌕 LiDAR

Connor Manning is a member of Hobu team, which is a group of five software engineers who have been at the forefront of open-source LiDAR software for fifteen years. They have been building open-source GIS and geospatial software for even longer.

Connor worked with the USGS Astrogeology team to build Moon Lidar. It visualizes the 6 billion point NASA LOLA using COPC (Cloud Optimized Point Cloud) and CesiumJS.

I'm not typing another word since you really need to see this for yourself.

FIN

I hope at least one of these sections managed to keep your mind off of IRL, even for just a few moments. ☮️