2024-07-12 –, For Loop (3.2)
LASDatasets.jl is a newly open-sourced Julia package that provides a central interface for manipulating LAS (LASer) data files in Julia. LAS is a file format that stores point clouds with different types of features (e.g. intensity, colour) and is widely used across industry and research. Join us as we discuss the LAS data format and how we have combined and built upon the best features of existing LAS processing Julia packages to give a central interface for all your LAS processing needs.
LiDAR technology is used to capture real-life assets as point clouds, which are collections of 3D points that have additional attributes such as intensity, colour, return index etc. The LAS file format is a system of storing these (often large) groups of points and their metadata in a structured way, and has a compressed format, LAZ, to increase storage capacity. LAS stores points according to one of a set of predefined “point formats”, which specifies the metadata you assign to each point and their data type. On the one hand, this makes sure your point cloud data is stored and typed properly, but on the other it makes the process of converting from Julia-native structures such as Tables or DataFrames quite difficult, since as a user you need to handle the data types yourself.
Although there are some great Julia packages that handle LAS data (such as LasIO.jl and LazIO.jl), there is no central entry point through which a user can simply read or write their LAS data to or from Julia without having to explicitly handle some of the data considerations themselves (e.g. whether the file is compressed, what point format it’s in, etc.). Furthermore, these packages don’t support all ten of the specified LAS Point Formats on offer. It is for this reason that we have further developed LAS.jl. We have used LAS.jl extensively in our LiDAR processing at Fugro, and so have ensured that it is both flexible and computationally efficient.
In this talk, we will discuss the high-level user interface of LASDatasets.jl and some examples of reading and writing data. We will then explore the underlying systems we use to parse Julia structures into correctly typed and formatted LAS data. Finally, we will outline some future opportunities for improving this package both in terms of user experience and computational performance.
You can find the code for the package here: https://github.com/fugro-oss/LASDatasets.jl
Mathematician and data scientist, interested in using Julia for high performance computing and cloud workflows
Dr. Megan Dawson is the Modelling Automation Team lead at Fugro, responsible for numerous cloud based HPC workflows efficiently processing large amounts of geo-data. With 15 years experience in industry her passion is problem solving to developing practical (and production-able) solutions.