PulseWavesIO.jl

Julia package for reading the PulseWaves lidar data format.

Overview

This is a native Julia implementation to read the full waveform lidar data format PulseWaves.

Installation

This package is not yet registered in the Julia registry, it needs to be installed from Github, for example with the code below.

using Pkg Pkg.add("url=https://github.com/danscr/PulseWavesIO.jl.git")

Example usage

The PulseWaves format is a pair of files, one with ending .pls, and one with .wvs. The load function takes the path to the .pls file, and throws an error if there is no .wvs file in the same directory with the same name (except the different file extension).

load returns the Pulse header, pulses, the (vestigial) waves header, the waveforms, the sampling type, and a single mandatory appended variable length record. The sampling type is not strictly part of the data, but it is returned as a convenience to distinguish system pulse samplings from return samplings. Unless you want to write PulseWaves files, you can ignore the waves header and the appended variable length record (in the example below they are ignored by using _ instead of naming the variable).

Example code: ``` Using PulseWavesIO, FileIO

header, pulses, wavesHeader, waves, appendedVariableLengthRecord = load("path/to/pls/file") ```

The pulse geometry (anchor and target XYZ) are read raw, and the function PulsesToRays( pulses, header) transforms coordinates into target values. The function returns an array of size Nx2x3, where N is the number of pulses, then the next dimension is anchor (1) or target (2) coordinates, and the last dimension is the Cartesian coordinates. The target coordinates are a direction vector with the length of one sampling unit.

The Waves contain a field 'DurationFromAnchor', which is simply multiplied by the direction vector of the ray to get to the starting point of the wave sampling. Each sample's position is offset a unit of the direction vector.

If the timestamps are needed, use the function RescalePulseTimestamps( pulses, header) to obtain GPS weektime data from the raw data. Note that there is a difference between GPS time and UTC of about 18-19 seconds, depending on what time your data was captured (search for "leap seconds" to learn more).

Experimental functionality

The folder dev contains two files named workflow*.jl. These were used to estimate plant area density from ALS data and go beyond the scope of reading PulseWaves files.

Roadmap

A goal of this project is to register the implementation for PulseWaves with FileIO.jl to make it available to users without any extra effort. Before that happens, the code needs to be thoroughly tested and refined to ensure it works smoothly. For now it was developed with Riegl Q680 data, so a bit finicky if you have other data.

Therefore, please report any issues you find or submit a pull request. Thanks!