Page 1 of 1

Quantifying erosion rates

Posted: Mon Dec 18, 2017 5:12 pm
by naeems9
Hello,

I am wondering if anybody can assist me on my dissertation project - coastal erosion.

My project is to compare 13 ALS LiDAR datasets obtained over a 6 year epoch; with a TLS point cloud obtained one month ago. Both datasets are in the OSGB36 coordinate reference frame.

I am wanting to quantify the rates of erosion. Erosion rates are high at the location I am using. Would it be possible to use the ICP surface matching algorithm on both datasets and obtain the distance the surface has moved to the reference surface. For example, use the TLS as the fixed reference and the LiDAR as the data?

This is a screenshot of the TLS data compared to the ALS from one epoch in CloudCompare.
Comparison CC screenshot.jpg
Comparison CC screenshot.jpg (237.07 KiB) Viewed 967 times

I greatly appreciate any help or guidance.

Kind Regards,
Naeem Stevens

Re: Quantifying erosion rates

Posted: Tue Dec 19, 2017 6:50 pm
by daniel
Hi,

I doubt using ICP on the whole cloud will work. As the deformation is probably not 'rigid'. If you are only interested in the vertical displacement for instance, you could use the M3C2 plugin, or more simply the 2.5D Volume computation tool (it's a side result).

Note: for the records we did try to implement a plugin in CC to do piece-wise ICP algorithm for this kind of application. However the results were not that concluding on the few cases we tested. This is a work in progress.