02-27-2017 02:47 PM
Good evening together!
Since my scanner is set up and waves are gone (i will write some experiences later) the alignment is a new issue.
I like the david software, the cleaning of scans, but the alignment could be improoved.
I went over to Meshlab, because meshlab allows piont by point alignment for difficult situations.
But Meshlab is very slow on my computer with a lot of scans loaded.
I tried geomagic design x and am very exited about the point by point alignment (automatic alignment is as bad as with david). And geomagic also offeres the ability to show how accurate your scans are aligned.
What i want to ask in this forum is the question: what kind of targets / registrations / software do you use for good alignment?
I use ball targets, but there are always scans, that cannot be aligned. And the accuracy of the good aligned single scans is around 1mm and higher deviation, which might be also improoved. Group alignment strategy works with geomagic and david and was used. So i dont ask for a software issue, i ask for a workflow :-) thank you!
Any ideas? Target stickers?
02-27-2017 03:14 PM
If scans are not aligning it may not be the alignment. Often distortion in the individula scans can give you scans that look close but are geometrically dissimilar enough to not give good alignment. This is especially true at the limits of the scan envelope.
02-27-2017 10:57 PM - edited 02-27-2017 10:58 PM
I also use geomagic for aligning and further reverse engeneering, but without any markers.
P.S. waiting for your waves solution!!!
02-28-2017 01:21 AM
Hi Micro, how can i reduce distortions?
The lens i have is a Fujinon cf16-HA1.
The setup is horizontal now and the camera looks at ~20deg to the scene.
Calibrated to 120mm.
Projector is tilted to some degree like in david setup.
camera sees a bit more environment than the projector patterns , but i think this should be ok?
It sounds possible to me, that there are distortions, because 2 software solutions cannot find a best fit...
Would it be a recommendation to delete scanned points being more far than the close focussed area scanned in the induvidual rotation?
waiting for the waves solution? :-) It is not as revolutionary as you might expect... i had vertical waves -->
get totally rid of all custom driver settings, such as anti aliasing and gamma and frequency fine adjustments,... to default, even if you think that you are in default mode.
I want to write a complete review with all my experiences and more details, please give me some time :-)
02-28-2017 02:00 AM
What kind of projector do you have? You should check keystone option, value should be 0, autokeystone should be off. That is the number 1 reason for bad alignment.
Original David Team Member
03-02-2017 01:11 AM
Optimizemulticameracalibration was enabled with single camera use.
In the old forum during beta testing of David 5 this was identified of beeing problematically.
I will disable this function for my next tryout.
//i dont know why, but i think that adjustments in the menue can change automatically, or are resetted unseen after crashing. I had unselected this before... and it was not the first time that settings were resetted by accident.
It would be perfect to have the ability to save all adjustments in a settings file, that can be reloaded. Especially for improovements on a stable basement (other hardware or fine adjustments).
keep you updated.
03-05-2017 01:28 AM
Hello, yesterday i tried out a calibration without multicamera optimization.
The alignment is still problematically and not accurate.
calibration panel is recognized well.
is it possible to have critical positioning in projector/camera angle(~20deg) or projector tilt angle or distance to object...?
What else can be the reason?