Learn LiDAR
19 articles on how LiDAR data actually works — written to be useful while you have a file open, not to rank for a definition.
Fundamentals
What LiDAR is, what a point cloud is, and the formats they arrive in.
- What is a LAZ file?What a LAZ file contains, which software opens it, and why some tools still refuse it.
- LAS vs LAZIdentical data, a tenth of the size. Which one to keep, and where COPC fits.
- What is a LAS file?Header, records and points: what is inside a LAS file, how the versions differ, and how to read one.
- What is COPC?How a Cloud Optimized Point Cloud is laid out, what it requires, how to make one, and what a host must do to serve it.
- What is an E57 file?The laser scan exchange format: scans with poses, station photographs, and how it differs from LAS.
- What is LiDAR?How a laser pulse becomes a measured point: returns, intensity, and why the positioning system sets the accuracy.
- What is a LiDAR point cloud?The data structure itself: how one point is stored, why LAS keeps coordinates as integers, and how a cloud differs from a raster or a mesh.
Elevation & terrain
Surface models, terrain models, and the difference that costs people money.
- DEM vs DSM vs DTMDSM keeps the buildings, DTM strips them, and DEM means different things to different organisations. How to tell which one you were given.
- How to create a DTM from LiDAR dataGround classification, filtering, rasterisation and interpolation — and what vegetation, bridges, water and steep terrain do to each step.
- How to create a DSM from LiDAR dataTake the highest point in every cell — and remove the noise first, because one bad return is now the answer.
Classification & analysis
Deciding what each point represents, and what to do with the labels.
- Point cloud statisticsWhat a classification histogram, return distribution, elevation percentiles and point density tell you about a file — and how to get them from the viewer.
- LAS classification codesEvery ASPRS class from 0 to 255, what changed between versions, and how to see what a file really contains.
- How LiDAR classification worksWhy a point can only be classified from its surroundings, why ground has to be solved first, and where every algorithm still fails.
- How to remove vegetation from a point cloudWhich classes to keep for each goal, and why filtering by exclusion leaves the vegetation you were trying to delete.
Formats & technical detail
Coordinate systems, records, point layouts — the bytes themselves.
- How it worksThe WebAssembly pipeline behind every tool here, and the limits it comes with.
- LAS point data record formatsFormats 0 to 10 byte by byte: what each carries, legacy versus LAS 1.4, and why waveform formats are rarely supported.
- What is a CRS in LiDAR data?Projected versus geographic coordinates, EPSG codes, and why the vertical datum causes more expensive mistakes than the horizontal one.
- LAS VLRs and EVLRs explainedWhere a LAS file keeps the coordinate system, the extra-byte definitions and the COPC index — and why conversion so often loses them.
- E57 vs LASWhich format to keep, and exactly what the conversion costs you — field by field, including the poses you cannot get back.
Where to start
Three routes through the same material, depending on what you arrived with.