andreaskrueck wrote: ↑Tue Jan 11, 2022 2:06 pm
I'm currently assessing multiple 3D indoor scanning options for a global client in the life science sector. The preferred solution would be an end-to-end vendor (scan to display 3D model), so there are some big players like Matterport, NavVis and FARO.
What I'm looking for now is any information on data strategy. Especially on those topics:
- Data cleaning: Do you need to keep the raw point cloud scan data if you already uploaded it into the vendors environments, resulting in a 3D model, that could be easily exported as point cloud again? And what are the reasons to keep them? Just as a backup doesn't make a lot of sense to me.
- Data storage/archiving: If the raw data is kept, I identified 3 options: Compressed files like LAZ, in a relational db system database like PostgreSQL or a NoSQL database, like MongoDB. Since the amount of data is quite large (multiple large sites globally), the last one makes more sense to me. Similar findings in this article: http://www.jait.us/uploadfile/2021/1029 ... 755987.pdf. Any experience you can share?
I'm happy about anything useful or any experience. Thank you in advance for your help
to answer your questions we need to look at how the data is getting processed and uploaded, and if 3D model means what you think it means.
Let's start with 3D model: A 3D model could be either a meshed point cloud (so you get an object file), or a complete BIM model (so every aspect of the model has attributes, like which type of window was scanned etc.)
None of the named vendors provides a BIM model, only a mesh or the visualized point cloud.
FARO would be presumably static scanning, where the registration is done in FARO Scene (FLS files are combined in a LSPROJ file) and then uploaded to FARO webshare (I guess). If I understand their concept correctly you have then the project visualized as a point cloud online, can move around there and measure.
You scan with SLAM (walking around rather than static scans), the data processing is done in NavVis Ivion (don't know if this is the processing software and or the visualization software) - in a cloud, you can download the result as a LAZ or E57. The "model" is meshed online, you can work around and measure distances and set tags. Very similar to google street view but with your data and possibly in buildings.
A static "scanner" (more like a camera with a depth sensor), the data processing is done online as well (as far as I know), but in this case (again I am not up to date on this one) your data now belongs to Matterport. Downloadable content is a XYZ file, online viewing is as well a mesh, I don't know if you are able to measure in there.
To answer your questions:
Data cleaning: I would always backup the raw data, because you can do so much more with it than with the processed and converted data.
I would use neither method. I would take the raw data, zip it according to projects and then store it whereever convenient (cloud, local server, USB drives). If the locations are relevant to every project zip, an index document, like location, dates scanned etc should be sufficient.
The LAZ way means your data was already completely processed where you will loose some information (more or less depending on choosen vendor. Example given: With the FARO data you will loose the structured scan positions, inclinometerdata, ..., with NavVis you will loose all panorama images taken.)
The database ways is just inserting it to a table, if you ever need them getting them out there might be a pain (don't know).
Other question for you:
Do you already have somebody who is going to do the scanning for you?
Because as far as I know the named vendors are only selling the hardware and providing the service to view your data online, but no actual leg work.