Reflector

This project produces 3D printable models from lidar data. Examples were generated from freely available data from Hong Kong but this should work for many lidar recordings taken aerially.

Overview

The point cloud is sampled into a height map, then interpolated to fill in any missing points before being output to a bin file for the viewer or final conversion to an STL or a normalized png for any height mapping software.

Instructions

First find some aerial lidar data in the LAS format. For my examples I sourced my data from here. Next, either view the data in the built in viewer for a preview or render it to a heightmap or STL file using the instructions below.

Creating a heightmap png for rendering in Blender

Run cargo run --release --bin converter -- --las-folder-path <PATH TO FOLDER CONTAINING LAS DATA> --pixels-per-unit-dim <use configured resolution per unit (x, _, z) square in the las file. 1 is a reasonable default> --rounds-of-interpolated-hole-filling 100 --consider-nearest-n-neighbors-for-interpolation 8 --output-path /tmp/test.png --base-depth 60.

Viewing LAS data in the viewer

First run cargo run --release --bin converter -- --las-folder-path <PATH TO FOLDER CONTAINING LAS DATA> --pixels-per-unit-dim <use configured resolution per unit (x, _, z) square in the las file. 1 is a reasonable default> --rounds-of-interpolated-hole-filling 100 --consider-nearest-n-neighbors-for-interpolation 8 --write-to-bin --output-path /tmp/test.bin --base-depth 60 to convert the LAS data to a bin file for viewing Next run RUST_LOG=viewer=info cargo run --release --bin viewer -- /tmp/test.bin.

Creating an STL

Run cargo run --release --bin converter -- --las-folder-path <PATH TO FOLDER CONTAINING LAS DATA> --pixels-per-unit-dim 1 --rounds-of-interpolated-hole-filling 100 --consider-nearest-n-neighbors-for-interpolation 2 --write-to-stl --output-path /tmp/test.stl --base-depth 60 to create /tmp/test.stl.

Example printed models

Example renders