-
Notifications
You must be signed in to change notification settings - Fork 84
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feature request: demystify point coordinates #253
Comments
Geo-referencing is a fairly deep topic. By the sound of it your data is in a known UTM zone (or similar) and you'd like to click on points and log those as WGS84 lat/lon? The EPSG code or WKT is available? I was looking just now at some RTK geo-referenced cloud in displaz. The z offset handling seems kinda wonky (IMHO) but that's easily fixed/changed. Potentially there is a problem if the offset is not chosen carefully by the LAZ writer, but I think an actual fix for that would be somewhat complicated. Have you tried using QGIS for georeferenced point cloud? It's pretty nice with the recent addition of COPC indexing. |
Yes, in this case the source data is Nad83/Conus Albers (EPSG 5070). This is parseable from the LAS header/metadata. My assumption was Displaz is already reading the source CRS and doing a transformation to UTM meters, since the point X,Y logged from this particular file (https://rockyweb.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/CO_SanLuisJuanMiguel_2020_D20/CO_SanLuisJuanMiguel_4_2020/LAZ/USGS_LPC_CO_SanLuisJuanMiguel_2020_D20_13S_BC_5409.laz) are in UTM meters. However, I've noticed the point X,Y logged from other files in the same LiDAR collection project (https://rockyweb.usgs.gov/vdelivery/Datasets/Staged/Elevation/LPC/Projects/CO_Southwest_NRCS_2018_D18/CO_Southwest_NRCS_B2_2018/LAZ/USGS_LPC_CO_Southwest_NRCS_2018_D18_w1014n1682.laz) are not in a CRS that I recognize offhand. In any case the source CRS should be found in the LAS header/metadata. I understand it may be defined in various parts of the LAS header/metadata, so implementation would need to consider at least a few scenarios. QGIS seems to have comprehensive tools for this type of exploration, but its enormous installation, horrendous UI, and inability to render the least bit of 3D without consuming every resource on my system are what drove me to look for alternatives. Displaz is hands-down the better tool for my use case (determining the clean prominence of mountain peaks by identifying the summit/saddle points). |
As far I know displaz is completely ignoring the CRS information so what it's displaying should match the raw numbers in whatever projection the data is in. In the case of EPSG:5070 it's meters with origin at latitude 23, longitude -96? The PROJ packages includes a cs2cs CLI tool for transforming between coordinate systems. It just needs the two relevant EPSG codes. Including PROJ, the EPSG database and the grid-shift files would be turning displaz into something more like QGIS. It's doable in principle, but it would not be cheap in terms of time or money. |
Yes, maybe there is a problem with the packaging of the first dataset then. Understandable--and if the required tools are not small, then it wouldn't make sense to grow the size of this application to the extent needed to support this feature. What might be incredibly useful instead is a summary of the LAS header fields/metadata that were read (in the log) after a file is loaded. Just knowing the CRS without having to inspect the LAS with additional utilities would significantly speed up the process of converting the raw values. |
So for that first LAZ file I pick a point at lasinfo has this to say about the projection:
It's EPSG:6342, rather than EPSG:5070 perhaps? |
With laspy reading the metadata I was getting 5070. Likely I was mistakenly reading one of the other project datasets while viewing the one which was already in UTM 13N. That makes a lot of sense, as the 2nd file is definitely not in UTM. |
I am working with geo LiDAR sets in LAS format, and my simple use case is that I want to easily georeference a selected point.
It took me forever to realize that Displaz is converting the source x,y coordinates to UTM meters.
This conversion makes sense for calculating vectors in meters. However, in order for me to figure out where the point is on the earth's surface, I have to know the UTM zone and then perform a conversion to a common coordinate system. It would be great if there were an option to show the point's lat/lon in a configurable notation (decimal degrees, UTM, etc), where applicable.
The text was updated successfully, but these errors were encountered: