You are not logged in. Please login or register.
SoliForum - 3D Printing Community → ATLAS 3D & FreeLSS → Turntable direction
My turntable runs counter clockwise and I am getting the exploded scans no matter what my settings or resolution are. Does the direction of rotation matter? If So Can it be changed in software (Motor direction Pin assignment?)
Counterclockwise is correct.
I don't know if reversing direction makes any difference, unless he (Hairu) is using that information in the reconstruction of the point cloud.
I can't seem to generate any sort of a useable scan.
Even setting thresholds to get a good test image fails. the grey suggests that to "Laser" threshold is incorrect. Have never evolves past the shattered scans. All Calibrated and so on, Wit's end. A document describing the settings variables with a bit more description for knuckleheads like me.
I can't seem to generate any sort of a useable scan.
Even setting thresholds to get a good test image fails. the grey suggests that to "Laser" threshold is incorrect. Have never evolves past the shattered scans. All Calibrated and so on, Wit's end. A document describing the settings variables with a bit more description for knuckleheads like me.
My first suggestion is to show us what the scan looks like.
Also show us your alignment so that we can have more information to possibly provide a more valid suggestion.
SIWEK,
Does your alignment image look like the picture I posted in your other thread? Please show us what your laser alignment looks like.
Uriah
Here are the pix
I may be wrong but I would think your calibration target should be centered on the table, not at the back edge. When the beams are on the object I would think they should appear as one beam. The issue is that you are getting two images super imposed due to the beams being spread apart. You can confirm this by covering one laser and run a scan with one and I bet you will not see your exploded issue.
You want it calibrated so beams converge at the center of the table I would think.
The calibration item is aligned to center of table. Budda is ALSO rotating about its center which is why there are two lines. Will give single scan a shot but then I don't see the purpose of two laser. The calibration images seem to match. Am I overthinking this?
In the picture you posted above it doesn't look like the calibration item is in the center of the table but it's hard to tell from that picture. There is a small hole in the center of the turntable. The front wall of the calibration item should be right at it. Your lasers should be going right over this hole.
Please open up the Buddha PLY file in MeshLab, then click the point cloud view icon, so we can see the point could only, then post that image.
@ Carl_m1968. The calbration looks fine, you can see the lasers lines converging in the middle of the object. The red lines you see on the edges are the laser projection on the backwall.
@Siwek. Is your image enlarged/zoomed in or, or it that the normal view you see when you look in camera mode? It appears to be zoomed in and cropped by quite a bit. The red cross lines, especially, look like what I see when I zoom into the image
Here is what I see.
Edit:
I figured out the reason why it looks like that. You're camera is set at the lowest resolution (640x480 video mode). You may want to try a scan at 5Mpixel (2592x1944 still mode), and see what you get. BTW, you should calibratethe lasers at that resolution before you scan.
The files below show pint clouds from the scan above. The image is a 5MP Camera view. eating a single laser pass run at full calibrated 
Attached two image results of hires, 1 laser scan. Wow didn't see THAT coming
As you can see from your point cloud the scanner sees the object (more or less) correctly. Once you turn off the mesh and look at the cloud only you can see what the scanner saw.
The problem is that the face (mesh) reconstruction is not very good, so it looks weird. The Raspberry Pi simply doesn't have enough horsepower to do a higher quality reconstruction.
For this reason, I don't even bother with the mesh that is generated inside the PLY, I only use the point cloud as the starting point, then I apply a few filters to do a better reconstruction, that is closer to what the object looks like.
Check this tutorial out:
http://www.soliforum.com/topic/12264/fr … -tutorial/
You should also do 2 separate scans, one with the left laser, then immediately after (without moving the model) scan with the right laser, then import both clouds into meshlab. Do they align closely?
I finally managed a full head scan of the Buddha. Turns out left vs right lasers were reversed. I need guidance on getting full height that doesn't cut off the top of the head
SoliForum - 3D Printing Community → ATLAS 3D & FreeLSS → Turntable direction
Powered by PunBB, supported by Informer Technologies, Inc.
