CloudCompare forum

Hi guys,

I am a newbie in both cc software and the forum, but I will be really appreciated if you can help me.

I am doing a thesis project about comparing photogrammetry and Lidar vertical accuracy in coastal monitoring. GPS data of three transects are taken as a reference (about 150 points each transect, knowing their easting northing and elevation). I need to firstly compare photogrammetry data to GPS data and then compare Lidar data to GPS data. Then I can analyze them to say which is more accurate to measure the elevation of sand dune.

To achieve this, I need first filter the data of vegetations(only the sand surface data is needed). Then, since the points of photogrammetry and Lidar are much more than GPS points. Thus I need to find all of them in one meter range distance to each GPS point and calculate the mean of them to compare with the relative GPS point, which is called 'interpolate'. Could you guys help me about that? I only have about one week time to do this. So I am really worried about that.


Thanks a lot.

To do what you want, you can either use the simple 'Cloud to Cloud distance' (with the GPS points as the compared cloud, and the lidar or photogrammetry clouds as reference). You can even activate the 'local modelling strategies' to interpolate the reference cloud (locally). BUT in this case you'll get the nearest neighbor distance, and not necessarily the vertical one:
http://www.cloudcompare.org/doc/wiki/in ... omputation

To get the vertical distances, as well as a more robust interpolation scheme and a lot of other advantages (like signed distances), you'll have to use the qM3C2 plugin:
http://www.cloudcompare.org/doc/wiki/in ... 2_(plugin)

However this one is harder to tame, so you should try to read at least the wiki, and maybe the scientific article or messages about it on this forum.

Thanks a lot for the quick response Daniel. You are really warm-hearted and it helps a lot.

I am still confused about filtering the LiDAR points. I would like to get elevation data of the bare sand. I don't want the laser pulse data reflected from vegetation. I want to take the ground hits (topology only) meaning the last returns from LiDAR. How to achieve this by using cc?

P.S. It does not need to be 100% completeness.

To filter the vegetation, you should look at the qCanupo and qCSF plugins:
http://www.cloudcompare.org/doc/wiki/in ... O_(plugin)
http://www.cloudcompare.org/doc/wiki/in ... F_(plugin)