Questionaire for the Q4 Update

Dear fam,

Thanks for trusting and using Revopoint 3D Scanners. We’re planning a significant update to our software in Q4. And we want your input because we want a better understanding of what we need to improve.
So we would be immensely thankful if you could take a few minutes to fill out the feedback form below.

Please comment and let us know if you have any more ideas not covered by the feedback form.

Many thanks,

Team Revopoint

I think it would be very good if there were an API for the scanner so that you can start and pause the scanner with your own software.

5 Likes

So as already mentioned an API so that the community can write their own software, that would promote the OpenSource idea very much and I suspect it would come out pretty cool projects. The hardware from you is actually top!

I think that comes Revopoint also to good if one has many user projects around it.

What still bothers me is the limited editing function in Revostudio, there is only one lasso, so you can’t draw straight lines along a contour to remove outliers. That’s why I always edit my point cloud in CloudCompare.

Also the denoising and alignment of several parts I do in CloudCompare meanwhile. Maybe you can learn something from it?

But CloudCompare is not really for this kind of scans, what bothers me is e.g. that I can’t select points that are in front of a contour, an example: cc cleaning - YouTube

So basically it all boils down to one thing: Revopoint should continue to sell its good hardware, focus on the firmware and write a good API, the software and oher cool projects (engineering with 3d printer) will then automatically come from the community.

1 Like

I hope you mean Revo Studio … CC is more than Revo Studio can offer , it is perfect for editing point cloud as it was exactly made for that kind of job and have all needed options .

Revo Studio have very long time processing one point cloud , imagine you have more than one , it gonna crash on itself . It is just not build to handle that kind of data and size .

Ah and for the question regarding merging Revo Scan with Revo Studio

Y e s ---- P l e a s e !

With the ability to remapping the textures after editing or merging .

2 Likes

Yeah CloudCompare ist a very nice Tool, but watch the Video…

especially in such cases you should be able to say that you only want to select the front points.

I am not really sure what is going on in this video here , as it seems the overlap points was not even cleaned .
And no you can’t select this kind of point with lasso as that would be impossible option , you need to use the proper function to clean it out before even start your merge .
I had worse case than that .
Check my tutorial in the Tutorial and Tips section of the forum , simple 2 moves to clean up quickly all bad points .

Or check my 2 videos on my YT channel related to it .

When I first started using the software this was my idea too. But now my workflow is scanning and fusing in RevoScan, and while it is processing import and work the fused files in RevoStudio. If this would be one program and it won’t be busy while it is fusing, then it would be good, but if you can not work on anything while it is fusing, it would mean less productivity.

1 Like

Hi @Martijn , you have a valid point here , sometimes I am doing other things in Revo Studio while scanning next things in Revo Scan , but basic functions would be welcome like texture remapping , small options to remove partial points before fusing etc… so I do not look into the whole one thing as that would delay a lot of stuff , merge and registration can take lots of time sometimes , so that would be good to have just the simple functions .

P.S I sent you email

Same direction as the API idea:

It would be cool to be able to point the scanner on a testing maschine and scan a new stage of the bending every time a trigger is send. (trigger could be send via mqtt)

so you could compare a part after different load cycles.

It’s a pretty special usecase, but abstracted, it could be useful for lots of people. In short, you could also call it “make revoscan scriptable”!

(And if you ask, I would prefer python)

1 Like

another feature could maybe increase scan quality:

beeing able to stop the table when a scan is performed, then stop the scanning, turn the table, stop the table and scan again.

This is how professional scanners often work.

1 Like

@basst that are per frame 3D scanners , portable scanners never stops as they relay on the surface feature for tracking and that make them no less professional, just different types. You can’t convert MINI or POP2 to have per frame scanning option , the hardware don’t have the proper optics for that job .

DATT is just a gimmick, there is not a special connection between it and the scanner beside the play button that start them both together .

Use the manual settings in Revo Scan for better results and quality of the scan or better use android app .

I know that the way it is now is not ideal but it going to change soon .

I see the difference between the approaches and of course, you can not compete against a Zeiss System.

But when I tried to scan a non moving object with a non moving scanner, the results of the measures where extremely good. As also shown by this guy: Revopoint POP2 Single View Accuracy and Precision Tests - YouTube

And in Marker Mode, there should not be a problem to stich those “single” Scans together, right?

It was not @basst , the results was wrong and I already proved it . Plus the algorithms changed so dramatically since then .

MINI is even worse on that than POP2 .
Lost of noises are induced and the accuracy is very bad , so keep moving your subject to stop accumulation of bad frames in one spot .

You never truly can capture single frame with a device that shot 10/frames per sec constantly and use Depth camera sensors to capture scanned object in motion .

Different technology is used to capture the 3D data .

I tried to do that many times , and I go nice scan results but lost very fine details in the process so not really what I was looking for , I could use denoiser at 10 and have the same results so totally pointless.

If that was an open source scanner and software I am sure our community members would transfer it into something amazing . But sadly it is not .

But your suggestion are great ! Don’t get me wrong … I wish they do that but I know they will never do it . The scanners are portable scanners not stationery .

Yes you can do it , but make sure you have not overlapped frames … as it can create some issues .
Marker mode don’t care about the object features , it cares only about the markers and it’s position .

But I would recommend you create a color marker if your object surface have less features as it going to be painful to merge .
I use blue dots and scan in color , this way my color Markers are visible on the point cloud after scanning what is easy to use as a marker point later while merging , especially on the edges so you know where to stitch it , other way it would be impossible , and using just feature mode for merging not always works with featureless objects .

The blue markets are only for the stitching references . You can use a sharpie as well if you can paint over the object surface .

Very interesting. Thanks a lot for sharing. Indeed, I had only tested it with the Pop 2 yet.

I’m very curios how this will evolve. Personally, I had my best results always with the turntable. By Hand, I mostly lost tracking or got some “smeared” models.

Open Source move from Revopoint would certainly be a revolution. Like the step from ultimaker with their slicer Cura.

1 Like

I have the mini. It scans decent, but it loses track and slides constantly not to mention that the software does not realign the scanned objects.

A lot of what I do is for business and it seems the information is a bit misleading regarding the size of objects that are trackable while scanning.

That said, the software needs many improvements.

…I would have rather paid $2500 for a usable product than a $1000 on a product that wastes $$$ in labor time trying to make it work.

Not blaming revopoint - I should have done more research…

What exactly you meant by that ?
What size objects are misleading?

You know that MINI was created to scan very small objects ? At the distance between 19-20cm with very small FOV …
So what is so misleading about that ?

What size of objects you trying to scan with , picture or screenshot would be helpful.

Here’s one instance of the size of items that we need to scan.
This particlar piece a few dozen times and it consistantly loses track. And I’m using the vanishing spray…

Again, Im not saying the scanner is junk, it simply was the wrong choice for my application - it scans small objects quite well…

Anyhow…

16657845096788577328035504735366|281x500