Skip to content

Latest commit

 

History

History
26 lines (17 loc) · 2.44 KB

README.md

File metadata and controls

26 lines (17 loc) · 2.44 KB

qrDAR

GPLv3 license made-with-python DOI

This is a method for using fiducial markers (akin to QR codes) to identify the location of tagged features in laser scans.

The examples here use AruCo markers attachted to trees, these are then scanned as part of normal scanning and their location and connected point cloud are extracted in post-processing.

Although the method has currently only been tested using a RIEGL VZ-400, with some simple modifications, it could be applicable to any intensity recording laser scanner e.g. not the ZEB-REVO as this only records xyz.

Preparing targets

Targets are printed on waterproof paper using a standard office laser jet printer. The paper has to be reasonably heavy weight (we used 365 gsm weight) so targets don't move in the wind or bend over - the flatter the target the better! Circular retro-reflective stickers are then attached to the markers to allow for automatic identification in post-processing. The paper and stickers are easily sourced online.

It is a good idea to locate markers where they are clearly visible (avoiding occlusion) and, if scanning a regular grid, facing towrads the centre of the plot.

A set of 250 markers can be found here here and a notebook to make your own here. A guide to deploying targets can be found here.

Scanning targets

There are no specific requirements for scanning codes as they should be adequately captured with normal scanning. They have been tested using angular resolutions of <= 0.04 where scans were done on a regular 10 m grid.

Post processing

Post processing is described in the find_markers.ipynb. This includes an option to extract the desired feature.