Skip to main content
Hello all, I am confronted with a task in which i have 100's of LAS files in which i need to find minimum distances from part of the data to another.

 

 

For example, in one LAS file i have trees, structures, powerlines water pipes etc (standard urban environment) is there a way i can find out the minimum distance between say the powerline and the nearest tree? OR adversely have a limit to say no trees within 5m of powerline and 10m water pipes and have FME spit out the a result set of trees/structures too close to the powerlines/water pipes etc?

 

 

 

any help would be appreciated.

 

cheers
Well, I have seen this done for extracting powerlines, etc, so it is possible. Having said that, it won't be an easy task. You'll probably need to use the PointCloudExpressionEvaluator in a way that lets you identify and isolate the individual features. How that's done, I'm afraid I'm not sure.

 

 

What I suggest is that you contact my colleague, Dmitri, who is an expert in handling point clouds within FME.

 

 

Just send us a support request (through this page) and mark it for the attention of Dmitri. And please do share here any useful info he supplies.

 

 

Regards

 

 

Mark

 

 

Mark Ireland

 

Product Evangelist

 

Safe Software Inc.
I will stick my neck out and reprhase your problem: The tricky part is to extract a useful set of urban features from LAS data. e. transform a noisy, unstructured point cloud into to a set of powerlines, buildings, threes and so on. And given the size of your data you don't want to do this by hand, but need some (semi)-automatic routine. 

 

 

I agree with Mark: If you haven't never done this before (I haven't, but I have former colleagues who have dived into the murky waters of LAS and automatic detection & classification) then you'd really appreciate the help of someone who has. 

 

 

Once you have a set of powerlines, a set of threes, a set of buildings etc. the rest of your analysis becomes much simpler, using standard tools as bufferanalys and so on. 

Reply