Transforming georeferenced map into default SCENE coordinate system

Discuss FARO hardware here.
Post Reply
LaserTrees
I have made <0 posts
I have made <0 posts
Posts: 3
Joined: Tue Jan 14, 2020 10:57 pm
4
Full Name: Brandon Alveshere
Company Details: University of Connecticut
Company Position Title: Student
Country: United States
Linkedin Profile: No

Transforming georeferenced map into default SCENE coordinate system

Post by LaserTrees »

This may be an unusual request, but has anyone found a way to transform files that use global or polar systems to the local system used in SCENE? I've found several references for transforming SCENE's local coordinate system to global systems, but nothing for going in the reverse direction. I'd like to transform a forest stem map (physical locations of ~200 trees; in UTM or polar w/distance and azimuth from origin points) into SCENE local coordinates, so I can align it with scans taken across the area. The reason I want to do this is, based on what I've read, sounds like transforming the point cloud out of SCENE local system requires survey points (scan campaign is long since over, so no way to determine where targets were by field survey) or identifying matching points between entity in global reference system and point cloud in SCENE local system (which is non-trivial in a natural forest setting). Each origin in the stem map corresponds to a different scan location (scanning and mapping done on a grid), so I essentially have a local relative polar coordinate system for every part of the stem map. Hoping based on this I could apply a scaling factor that would essentially convert the true measured distances/azimuths from the stem map to the local system used in SCENE. I have talked to FARO support and they couldn't tell me anything about local system they use (seriously), so was hoping someone on here might be able to chime in. Any insight into addressing this issue directly or through a workaround greatly appreciated.

Kind thanks,
Brandon
jedfrechette
V.I.P Member
V.I.P Member
Posts: 1236
Joined: Mon Jan 04, 2010 7:51 pm
14
Full Name: Jed Frechette
Company Details: Lidar Guys
Company Position Title: CEO and Lidar Supervisor
Country: USA
Linkedin Profile: Yes
Location: Albuquerque, NM
Has thanked: 62 times
Been thanked: 219 times
Contact:

Re: Transforming georeferenced map into default SCENE coordinate system

Post by jedfrechette »

If the origin of each stem map is the same physical location as a scan position than, at least in principle, this is pretty easy. If the scanner's compass was enabled during scanning and registration than your scans should already be oriented to North. All you need to do is lookup the position of each scan (you can find this in the scan properties inside SCENE) and and use that as the origin of your stem map. Once you have the origin you can calculate the stem positions relative to that using your distance and azimuth measurements.

In practice, this isn't going to work perfectly because of errors in your azimuth measurements and the scanners compass measurements. You'll probably need to calculate an additional rotation to apply to each stem map so that it lines up with the trunk positions in your scans. Depending on how accurate you need this to be you may be able to just eyeball the rotation manually or you may want to do an actual best fit adjustment, which would likely require marking reference points in your scans.
Jed
LaserTrees
I have made <0 posts
I have made <0 posts
Posts: 3
Joined: Tue Jan 14, 2020 10:57 pm
4
Full Name: Brandon Alveshere
Company Details: University of Connecticut
Company Position Title: Student
Country: United States
Linkedin Profile: No

Re: Transforming georeferenced map into default SCENE coordinate system

Post by LaserTrees »

Hi Jed,

Thanks for the quick response; the method you suggested worked well for producing an accurate plot grid that aligns with the point cloud (didn't see the local coordinates under properties before because I was looking under "sensors"). Unfortunately the calculation of the stem map positions is still not quite right, but I think it may have something to do with the negative Y coordinates I get from the scan properties and the trig-based approach I'm using to calculate stem positions from known origin coordinates. At this point it is still unclear to me why some Y coordinates are positive and others are negative, given their relative positions on the scanning grid... Assuming the scanned plots were oriented North, would all the features identified to the North and East of the true plot origin (SW corner of plot) be positive? Just trying to wrap my head around this local coordinate system...

Thanks!
Brandon
jedfrechette
V.I.P Member
V.I.P Member
Posts: 1236
Joined: Mon Jan 04, 2010 7:51 pm
14
Full Name: Jed Frechette
Company Details: Lidar Guys
Company Position Title: CEO and Lidar Supervisor
Country: USA
Linkedin Profile: Yes
Location: Albuquerque, NM
Has thanked: 62 times
Been thanked: 219 times
Contact:

Re: Transforming georeferenced map into default SCENE coordinate system

Post by jedfrechette »

LaserTrees wrote: Sat Mar 14, 2020 11:57 pm Assuming the scanned plots were oriented North, would all the features identified to the North and East of the true plot origin (SW corner of plot) be positive?
I'm afraid I don't quite follow you here.

If you're working with the scanner coordinates they are a right-handed Cartesian coordinate system. +z is up, +y is north, and +x is east (assuming the scanner's compass was used). The origin of that global coordinate system will be at some x, y, and z position depending on which of the scanner's sensors were enabled and how you did your registration. If you have multiple scans and they have been aligned to each other then they will all be in that same global coordinate system.

If you want to back out each scan's individual local coordinate system the way I would approach it is by exporting your scans to either .ptx or .e57 format. Either format will store the scan's point coordinates in the local coordinate system as well as the transform necessary to align them to the world coordinate system. I'm sure you could do the same by inspecting the properties of the scans inside SCENE, but I would find it easier to directly access either the 4x4 matrix from the ptx file or the translation + quaternion from the e57.
Jed
Post Reply

Return to “FARO”