I am now sure this is a program glitch that needs repair. I spent some hours messing with different point clouds and different projects, and rolled back TBC on a different computer. What I found was the glitch was introduced with the refinement option of the point cloud region classification. Prior to the refinement tool, the recommend refinement approach was to classify the ground region from the first processed ground region. Now, when you run the point cloud classification tool with one step at a time, first classification, then refinement, the mixing of the scans and separated point cloud region happens only at the refinement step. No unintended mixing of scans and regions happens in the initial region classification step. The error is in the refinement step.
Trimble folks that should monitor these communities, it would be great if you could confirm this.
Original Message:
Sent: 06-03-2025 10:13
From: ian bissonnette
Subject: TBC Region Classification Not Respecting Scan Individuality
yes exactly. don t classify in the master if that is an issue.
one other question did come to mind. there is no way that you have multiple scans highlighted first and then you disable those unwanted ones and it is sticky that way.
more likely you turn the unwanted scans off first then select the ones you want to classify. but i could see how the first way would likely be problematic as tbc seems to hold selections awkwardly.
it does seem like a bug to me either way.
micheal mentioned that tbc will do more than one adjustment. i m interested to know more about this. we are stuck on 2023 till we decide that there are improvements worth paying for and this might be one of them. i am mentioning this here because it would be good to know if there is a way to "lock" datasets after an adjustment or in this case after a classification. this would be powerful.
Original Message:
Sent: 6/3/2025 12:37:00 PM
From: Clayton Bradshaw
Subject: RE: TBC Region Classification Not Respecting Scan Individuality
Hi Ian,
Thank you for the continued response. Did you see the PDF I attached that shows that what I am saying is actually happening?
I wish Trimble would consider this and possibly take the time to respond. We also use Agisoft Metashape and when I post a complex thread on their message board, they have folks that respond promptly when needed and appropriate.
I can't tell if this is a major bug, or I am doing something wrong. It just seems so obvious that the software shouldn't behave this way. But I may be wrong.
I appreciate the ideas you share. I may try the export/import of LAZ files into a master, but I would expect the same thing to happen. I guess the key would be in the master you could never classify regions.
Thank you again for your input.
CB
------------------------------
Clayton Bradshaw
Original Message:
Sent: 06-02-2025 14:02
From: ian bissonnette
Subject: TBC Region Classification Not Respecting Scan Individuality
Clayton, I agree that if you are seeing the software compute regions/scans that it should not be then this is an issue. And I agree that it would be nice to have all data in one tbc file. But I also see the software has some funny things happen in point clouds. Often the ribbon shows region color but really it is showing true color. The properties don't switch and you have to right click and show properties each time.
TBC will also not adjust two networks in one file which is really annoying. Maybe this is one of your hesitations when you were talking about sx10 data. A way around this is to do one adjustment if possible and then import control points as a csv into another data set(another TBC file), but this is not always a good solution and can get pretty messy if you have many days and a progressive control network.
I guess another option for how you are trying to work would be to adjust/classify your day/part of a day data in one TBC file then export the LAZ for each of these. and bring that classified LAZ into the master file. I realize this is a workaround, but if you need to move forwards this could be an option.
Original Message:
Sent: 6/2/2025 2:48:00 PM
From: Clayton Bradshaw
Subject: RE: TBC Region Classification Not Respecting Scan Individuality
Hey Everyone. I wrote up a more thorough summary in an email to our vendor support team and wanted to share that more thorough summary here. I can't believe this isn't a bigger deal for everyone. I feel like I must be using TBC incorrectly or so differently than the rest of the group. This is a major issue for our work flow and best practice and operation of the program. We will have to think through quite a bit of revisions to make this work going forward. We haven't used the reference projects, or what Mike and Ian suggest. We want all of our data in one master clearinghouse as often as we can. But maybe the reference project tool is powerful and the only way to be using the program. I wasn't aware you couldn't include multiple days of data in the same TBC file. That seems a little odd though and it seems like a lot of TBC files. Regardless, Here is the right up it is attached.
------------------------------
Clayton Bradshaw
Original Message:
Sent: 06-01-2025 14:17
From: Clayton Bradshaw
Subject: TBC Region Classification Not Respecting Scan Individuality
Thank you Mike and Thank you Ian for responding. I appreciate your input. The issue with this approach for our work flow as related to survey data collection is that Trimble Business Center is being used as our Master clearinghouse and storage of our measurement data. With multiple different UAV flights the approach to having multiple different TBC files is not problematic. The issue is with point cloud information coming from SX10 type scanning, maybe X7 type scanning too (I am not sure if the different TBC files would be problematic for X7 as we hardly use the tool, it did not prove to be a very valuable tool for our use). The SX10, though, and scanning to point cloud with it, is used all the time, daily, multiple times a day in some instances.
To not be able to process each instance and site conditions with that tool would not align with our approach. We need all of our measurement data in one TBC file, which means we would have a mix of site conditions over time for those scans collected with that tool.
I don't understand the limitation or process we are having issues with TBC. For TBC to classify a scan or point cloud into regions, you must first select the target for process consideration. Why is TBC allowing and including more into the process consideration than what is selected. Why, during the process, is it forcing scans from multiple different site conditions, or days of measurements, or from different UAV point clouds, to be processed at the same time and mixed, so that when I select only point cloud data, let's say, from May 20, 2025, is it also considering and mixing point cloud data from June 14, 2024?
I would like Trimble to answer this. It seems like a program bug. It didn't used to operate this way and it doesn't mix up all the scans, the scans and regions it mixed up tend to appear random.
Thank you.
------------------------------
Clayton Bradshaw
Original Message:
Sent: 05-31-2025 11:53
From: Mike Tartaglia
Subject: TBC Region Classification Not Respecting Scan Individuality
Here is the workflow I have our clients use.
-Create a master file with the design surface
- Create a template with just the calibration
- Create a file for each drone flight, do the cleanup, breakout into regions. Then create the needed surface
- Use reference file tool in the main project and use the surface from that drone flight file to compare quantities against.
- You then use this tool to run cross section to see the progress between the progress flight.
this keeps everything thing neat and manageable.
Hope that helps.
Mike
------------------------------
Mike Tartaglia
Original Message:
Sent: 05-31-2025 07:26
From: ian bissonnette
Subject: TBC Region Classification Not Respecting Scan Individuality
clayton. have you tried or considered having multiple tbc projects. a master and then one for each scan day. classify each day separate that way and then reference them in the master.
i only recently saw/tried this in tbc so do not know the limitations in your case. but i do know that renaming all those scans in tbc is real annoying and slow and the properties do not auto show up like on most other objects.
------------------------------
ian bissonnette
Original Message:
Sent: 05-30-2025 15:24
From: Clayton Bradshaw
Subject: TBC Region Classification Not Respecting Scan Individuality
Hello,
I have a project that has many point clouds, over several different time stamps and with each different time the site is different because of project grading.
As a process practice, we bring in point clouds, whether by UAV, LiDAR or terrerstrial scanning and then rename each "scan" to be prefaced by the date of the point cloud data. By "scan" I mean what TBC groups as "scans". This includes all the previous mentioned types.
After import of each point cloud we also rename or move all points in that scan to a unique dated default point group classification.
After that step, we would classify point cloud scan into regions and rename each region that is classified with the preface of the date of capture.
So for example,
Import UAV point cloud file "UAV Flight 3", then
Rename the name TBC gives to that "Scan" to 25-05-28_UAV Flight 3.
Select all of the points from this scan and create a new region can 25-05-28_UAV Flight 3 Default.
Then select only the point that are in that region named 25-05-28_UAV Flight 3 Default and classify and refine.
Then rename each region with a preface of "25-05-28_UAV Flight 3", i.e. 25-05-28_UAV Flight 3 Buildings, etc.
This kept all points separated by when they were captured based on site conditions. You wouldn't want TBC to evaluate the ground region and consider point cloud points from multiple different days and multiple different site conditions. So this is how we work planned to keep things separate and as far as we know it was working. Until the latest release.
Now, for some reason, TBC is allowing other scans and points cloud regions to be analyzed with the new data that has entered the project. So now, when I have looking at 25-05-28_UAV Flight 3 Ground-Refined, some how points from 24-03-27_UAV Flight 2 "scan" are also in this refined group.
It shouldn't be possible. Everything but what I want to consider for classification is turned off and isn't selected, but somehow, when the classification is done, points from out dated data have affected the analysis for current point classifications and are migrated into these point classifications. Luckily they remain in the correct scan, so I can start all the way over and do it again.
I don't understand what happened. Can someone please help? Please.
------------------------------
Clayton Bradshaw
------------------------------