Help with scanning with David SLS-3

Write here about construction, ideas, equipment, tips n tricks etc. related to structured light scanning
Post Reply
thanksleo888
Posts: 2
Joined: 13 Jan 2019, 21:31

Help with scanning with David SLS-3

Post by thanksleo888 » 13 Jan 2019, 23:42

Hi All,

New to this, I have a piece of a front guard I would like to scan.
I have started to scan the guard with David SLS-3 but the software is struggling to align the scan once its bigger that 250mm. I am using the 240mm calibration screen


The part size is 400mm by 500mm see the ruler in the photos for reference

Now for the noob question:

How can I scan this guard without getting alignment issues?

Do I need to adjust the Keystone on the projector to get a better scan area?

What would be the best orientation to scan such a part?

Would I need to scan both sides to get a solid model?

Scanner is David SLS-3 software 4.5.3

Thanks in advance
Leo
Attachments
IMG_4638.JPG
IMG_4637.JPG
IMG_4636.JPG

User avatar
Micr0
Posts: 333
Joined: 15 Nov 2016, 15:20

Re: Help with scanning with David SLS-3

Post by Micr0 » 18 Jan 2019, 01:20

thanksleo888 wrote:
13 Jan 2019, 23:42
Hi All,

New to this, I have a piece of a front guard I would like to scan.
I have started to scan the guard with David SLS-3 but the software is struggling to align the scan once its bigger that 250mm. I am using the 240mm calibration screen


The part size is 400mm by 500mm see the ruler in the photos for reference

Now for the noob question:

How can I scan this guard without getting alignment issues?

Do I need to adjust the Keystone on the projector to get a better scan area?

What would be the best orientation to scan such a part?

Would I need to scan both sides to get a solid model?

Scanner is David SLS-3 software 4.5.3

Thanks in advance
Leo

For someone new to this, you picked a fairly advanced part to start with. Parts with minimal thickness can be difficult to scan correctly. Likewise parts with large areas of relatively featureless topology can be difficult to align if scanned in pieces. I would suggest using some form of markers that can help you align scans.

First, The keystoning on you projector should be completely off.

You are going to have to take a good number of scans to get that entire thing. You should experiment with it until you understand the process. A can just tell you how to scan that. start by scanning some smaller objects. one that fit easily into the field of view of you scanner. When you have a feel for it try bigger parts.
µ

thanksleo888
Posts: 2
Joined: 13 Jan 2019, 21:31

Re: Help with scanning with David SLS-3

Post by thanksleo888 » 28 Jan 2019, 11:22

Hi Micr0,

Thanks for the reply

I was able to scan the guard by using a pencil and drawing some marks on the guard then using the "Contact pair selection" to align each scan.
I have found the shape of the guard is not focus on the edges of the scans

Another issue is when I align some scans they create a edge within the scan

I am scanning this with the 240mm scale would it be better to come down to 120mm add more marks and scan smaller sections?

Thanks in advance

Leo
Attachments
2019-01-28 22_10_06-.png
Scan not in focus
2019-01-28 21_59_37-.png
edge within the scan
2019-01-28 21_58_42-.png
edge within the scan
2019-01-28 21_52_26-.png
all scans

User avatar
Micr0
Posts: 333
Joined: 15 Nov 2016, 15:20

Re: Help with scanning with David SLS-3

Post by Micr0 » 30 Jan 2019, 13:34

thanksleo888 wrote:
28 Jan 2019, 11:22
Hi Micr0,

Thanks for the reply

I was able to scan the guard by using a pencil and drawing some marks on the guard then using the "Contact pair selection" to align each scan.
I have found the shape of the guard is not focus on the edges of the scans

Another issue is when I align some scans they create a edge within the scan

I am scanning this with the 240mm scale would it be better to come down to 120mm add more marks and scan smaller sections?

Thanks in advance

Leo
Again, very thin parts are difficult to scan well. The problem is two fold. 1, the scanner tends not to like hard edges and 2 (more importantly), you have very little in the way of features to register the thickness and align the front and back surfaces.

Part of what is happening in your case, I suspect, is that the errors in calibration are being revealed by the very thin part. Calibration is never perfect and the error shows up as scans that don't perfectly mesh together. If you want to see examples of scans that really mesh well look at Narmella's work on the old forum. He was using DAVID with the still capture function and an SLR and was able to get some amazing results. You can also check the amount of error in you calibration process by watching the debug console during your calibration process. Do a search on this forum for BigBomber's discussion on the subject. That will help you immensely.'s a trick I developed when scanning some sheet metal covers for motorcycles. I take grey plasticine modeling clay and and make quarter size balls that can be stuck to the edges in a few places. This give these edges some volume off axis for the scans to register to. When you are done scanning the part, get some scans of those areas with the clay removed to that in editing you can delete the clay bits and patch with the other scans.

Now, Here
µ

Post Reply