Detect identical points shared by two different clouds

Feel free to ask any question here
Post Reply
carlisle_h
Posts: 3
Joined: Mon Apr 18, 2016 8:50 pm

Detect identical points shared by two different clouds

Post by carlisle_h »

I'm looking for a way to compare two separate point clouds in order to determine if any points are common to both. I have two files, let's say a large and a small one, and I suspect the small one is simply a subset of the larger one, but I'd like to see if that is true for the entire dataset. I have found a few ways to identify duplicate points within a single cloud, but not comparing two. At first I thought I could just merge the large and small together, and then identify duplicate points, but that does not tell me which ones were common to both.

thanks,
Carlisle
daniel
Site Admin
Posts: 7710
Joined: Wed Oct 13, 2010 7:34 am
Location: Grenoble, France
Contact:

Re: Detect identical points shared by two different clouds

Post by daniel »

You can simple compute the C2C distances between the two clouds (use the small one as the 'compared' cloud).

Then you'll see if all points have a 0 distance (or almost zero). And you can even segment the cloud based on the distances to know how many points have exactly a zero distance.
Daniel, CloudCompare admin
carlisle_h
Posts: 3
Joined: Mon Apr 18, 2016 8:50 pm

Re: Detect identical points shared by two different clouds

Post by carlisle_h »

Thanks Daniel, that is very useful. I had not used the min/max tool before, very cool.

One follow up question that this brings up. It seems I don't have any distances that are actually zero, but there are thousands which are very small values, like 0.00030033. At what threshold should I consider these values to effectively be zero?
daniel
Site Admin
Posts: 7710
Joined: Wed Oct 13, 2010 7:34 am
Location: Grenoble, France
Contact:

Re: Detect identical points shared by two different clouds

Post by daniel »

It mostly depends on the file/source of the data. For instance If it comes from an ASCII file with a few digits (or very big coordinates) then you may be already at the numerical accuracy limit.

As a rule of thumb you can consider that the smallest value that can be properly represented in memory is about the biggest coordinate in your cloud divided by 10^7.
Daniel, CloudCompare admin
lna2020
Posts: 1
Joined: Mon May 09, 2016 12:29 am

Can I key in the coordinate to find the point?

Post by lna2020 »

I have (x,y,z) ,and I want to fine where is this point in CC.
Can I key in the coordinate to find the point?
What should I do?Thanks.
daniel
Site Admin
Posts: 7710
Joined: Wed Oct 13, 2010 7:34 am
Location: Grenoble, France
Contact:

Re: Detect identical points shared by two different clouds

Post by daniel »

The best way to localize a point is to create a sphere (with the Primitive Factory) and then translate this sphere with 'Edit > Apply transformation' (use the second tab, and simply set the points coordinates as 'translation').

You'll have to set a small radius for the sphere if you want to properly localize the point (you can change it afterwards in the sphere properties).
Daniel, CloudCompare admin
Post Reply