Dear DAVID Community,

As you are probably aware, the entire DAVID company, team and product range has been integrated into HP Inc.
This forum is now read-only. Relevant content has been migrated to HP and merged into HP's Support Forums on November 1, 2016.

To start new discussions for 3D Scanning, please register and post your new topic at the HP Support Forums

Man vs Machine

Which camera works? Where can you buy an adequate laser? And more...

Re: Man vs Machine

Postby BJ64 » Sat Jun 06, 2009 1:06 pm

Just another idea - would freezing it help to keep its form? Maybe even some liquid nitrogen - but that might make it extremely brittle I guess. Perhaps something else along those lines...?

I reckon the super glue might still be a cheap experimental avenue...

BJ :)
BJ64
 
Posts: 118
Joined: Mon May 11, 2009 1:02 pm
Location: Melbourne, Australia

Re: Man vs Machine

Postby Bongobat » Sat Jun 06, 2009 5:32 pm

With a little more patience...

EDIT: Can I get a gallery nod for this one now? Or do I have to scan bacteria...
Attachments
wasp2.jpg
New scan of Wasp
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Re: Man vs Machine

Postby WalterMo » Sat Jun 06, 2009 8:04 pm

Hi Greg,

Great result. Unbelievable! The thin legs are clearly visible. I can understand what patience was necessary for the preparation of this small insect. Hats off!
May I bestow you the title as the master in DAVID microscanning.
But one question: Have you used a software to „correct“ some scan imperfections? I mean out of DAVID.

Walter
User avatar
WalterMo
Moderator
 
Posts: 2363
Joined: Mon Apr 02, 2007 6:52 pm
Location: Braunschweig, Germany

Re: Man vs Machine

Postby Bongobat » Sat Jun 06, 2009 8:54 pm

Hi Walter,

Yes I couldnt help but cheat a little bit.

The first thing I did was take a small paint brush and some acetone and clean off the caked up paint from the wasp and lightly respray those areas. The paint still caked a bit or became fuzzy if too light a spray was used but it was better. Unfortunately Only after the scan did I noticed that I moved his upper set of hands closer to his body durring the cleaning. I scanned this one with a slower sweep and many more angles. The more angles helped with the wings but not much. Shapefusion has a hard time with thin planer objects. It eats them away and puts holes in them. It likes 3D not 2D points. I used a zbrush tool to drag them out to the edges and I also used the smooth tool to smooth some of the paint problems. No detail was added though if anything I accidentally erased some. I did absolutely nothing to the legs though. They are exactly how they were scanned.
Attachments
touchup.jpg
small touch up to bring the wings back
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Re: Man vs Machine

Postby BJ64 » Sun Jun 07, 2009 3:32 am

Truly amazing work there Greg :P

Although, if you do happen to scan in a Bacteria, I think you'd be in the running for some sort of Nobel Prize.... LoL

Enjoying your posts,

BJ :)
BJ64
 
Posts: 118
Joined: Mon May 11, 2009 1:02 pm
Location: Melbourne, Australia

Re: Man vs Machine

Postby Ordibble Plop » Fri Jun 12, 2009 12:21 am

Fine scans in all ways.

For small objects like this, a technique miniatures painters use to undercoat models is to use a gesso primer (commonly Liquitex gesso - http://www.liquitex.com/Products/surfprepgesso.cfm )

It goes on with a paint brush quite globby but shrinks to form a thin skin. You can thin it out with water (I wonder if dipping in a thin solution would work for something like an insect) and also with matte extenders.
Ordibble Plop
 
Posts: 103
Joined: Mon Mar 12, 2007 7:18 am
Location: New Zealand

Re: Man vs Machine

Postby Bongobat » Fri Jun 12, 2009 1:00 am

Hi,

I have asked the question on how to prepare a insect for scanning at insectnet.com: http://www.insectnet.com/dcforum/DCForumID1/1979.html

@ Dave
Thanks, for the info on the gesso. That looks like it may be a good alternative to the spray paint.

Also I have made my first kill :o Another paper wasp or to be more exact a polistes fuscatus. I hope to get even better results with this specimen 8)

Greg
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Re: Man vs Machine

Postby MagWeb » Fri Jun 12, 2009 9:17 am

Hi Greg,
read your thread at insectnet.com and took a short look at this paper about bleaching insects.

I´m not sure whether using H2O2 will give a useful result for scanning. I´m using this in my workshop to bleach wood.

H202 is a rather aggressive bleach. (Never use some brush or pot with metal >>> H2O2 in contact with metal is going to boil after a short time)
For some purposes I´ve to use small hair brushes which loose their hair very quickly...
So, if you give it a try start low concentrated and be patient.

But the either negative points of bleaching:
- You need ammonia to activate the H2O2 (wear a gasmask!). Ammonia will, besides activating H2O2, also splits the insects cells - so the chitin might swell up?
- And bleaching will give a transparent not opaque white object....

Another option to bleach could be oxalic acid... I´ll do some tests for you

@ using gesso: Think gesso is made to be a filler. Means the paint contains materials (e.g. chalk) to fill cavities of a material to give a smooth , closed surface. This is, what we allways try to avoid concerning scanning. I think a paint with little filling substances as possible should be better.
(there are some white colored wood stains - stains try not to fill the pores.... let´s see)

I´m going to hunt some bugs

Gunter
MagWeb
Moderator
 
Posts: 2499
Joined: Wed Jul 18, 2007 8:48 pm

Re: Man vs Machine

Postby Bongobat » Fri Jun 12, 2009 9:38 pm

Hi Gunter,

MagWeb wrote:But the either negative points of bleaching:
- You need ammonia to activate the H2O2 (wear a gasmask!). Ammonia will, besides activating H2O2, also splits the insects cells - so the chitin might swell up?
- And bleaching will give a transparent not opaque white object....

I think you are right. I read the paper better today and it seems the H2O2 will make the exoskeleton clear but will also make the organs swell up and be misplaced so they use a clearing solutions to clear the organs and create a glass like outer form. I think I read that right? Anyway a clear exoskeleton is not good for scanning. I had the thought after bleaching for the clear look you could then use a food coloring or some sort of dye to add back the color? Seems a little over the top though. Their picture of the ant was bleached for 3 days with 35% H2O2 but I dont think they mentioned an activiator like amonia? They did tests with light exposure though. I dont think I have that much patience.

MagWeb wrote:Another option to bleach could be oxalic acid... I´ll do some tests for you

Thanks :D

MagWeb wrote:@ using gesso: Think gesso is made to be a filler. Means the paint contains materials (e.g. chalk) to fill cavities of a material to give a smooth , closed surface. This is, what we allways try to avoid concerning scanning. I think a paint with little filling substances as possible should be better.
(there are some white colored wood stains - stains try not to fill the pores.... let´s see)

So yes back to some sort of coating or Dye. The gesso if it works as Dave said and shrinks may work well or a wood stain sounds possible too. On the other forum someone mentioned tippex so I have a few more options now thanks.

MagWeb wrote:I´m going to hunt some bugs

Happy hunting :wink: Dont attempt to catch anything bigger than yourself.

Greg
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Re: Man vs Machine

Postby Bongobat » Sat Jun 13, 2009 12:10 am

Another thought :idea:

What would happen if I used fluorescent ink to coat the wasp? I then scan it with a UV LED or a small halogen UV lamp in my projector? Can I get a thin line like this? I should also be able to get texture shots of the insect as the fluorescent ink is invisible under normal lights.

Image

Greg
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Re: Man vs Machine

Postby Bongobat » Thu Jan 14, 2010 7:54 pm

Hi,

Just an update.

I have learned from these planeless scanning tests a couple of important things. 1. The large intial wavinesss problems in the non remmembered plane pose scans (UseRememberedPlanePoses = False) problably were caused by the current limiting settings of the easydriver. The easydriver version I use before had a wacky loose pot that would move and may have been giving me inconsistant current levels to the motor. It seems when the remembered plane pose option came along it eliminated the problem and allowed the system to have good scans and 2. Perhaps my initial way of attaching the stepper motor to the worm shaft was not as perfect as needed. Even though planeless canning was possible with remembered plane poses option I continued to work on the problem

So since my last post here I have been sourcing new hardware parts to make my system better. I have now put together a system made of very cheap and readily available parts that is capable of planeless scanning with the UseRememberedPlanePoses option set to false - Meaning this scan is relying completely on the accuracy of the hardware mechanics and electronics. The computer is not compensating for mistakes. Here is one of my initial tests using the new system with the $4 deal extreme line laser and the SPC900NC webcam: A black baseball cap. Result is without smoothing and interpolate at 1.

Soon I will start a thread explaining the new system

EDIT: Added scan of paper towel roll with perforated pattern.

Greg
Attachments
planelesshat.jpg
Planeless test without using remembered plane poses
planelesspapertowelroll.jpg
Planeless test without using remembered plane poses
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Re: Man vs Machine

Postby RAYA » Thu Jan 14, 2010 8:07 pm

Hi Greg

Waiting for such, very cheap and readily available parts, is not easy , may you name the thread ,without stepper motor ,I first tried planless scanning with a kitchen timer and put laser line out of webcam view at start.

Ray
User avatar
RAYA
 
Posts: 446
Joined: Wed Oct 08, 2008 5:40 am

Postby hal » Thu Jan 14, 2010 11:20 pm

Hi Greg,

cool scans 8)
Ok, now you achieves good detailed results, but what tell me about the problem of little/big deformations?
With my setup for planless scans, if I forget to be VERY precise in all steps, I can obtain big problems: portions of the scan with noise and/or good single scans, but that reveals a deformation after the alignment. And just a little deformation, if multiply for a lot of scans, became a big problem.
Actually I can't to achieves, with planeless scans, same good results as with the panels.
So I ask you if you have find a good way to jump this problem, too.

Please keep up your great work up,
Mattia
User avatar
hal
Moderator
 
Posts: 1527
Joined: Thu Mar 01, 2007 4:08 pm
Location: Cavaria con Premezzo (VA), Italy.

Re: Man vs Machine

Postby Bongobat » Fri Jan 15, 2010 12:21 am

Hi Mattia,

These are just my first tests with a new system and they are clearly better results than my old system but are they perfect? NO, problaby not. I am striving for results like the kind that can be obtained by Walter's impressive scanner and I think I am just a little closer now. I belive Walter's scanner sets the bar to which all other systems should be measured. A very high bar :wink:

However here are the problems I noticed from my tests and it would be helpful to me if you could tell me if you have the same problems with your system.

1. The faster the scanning speeds the larger the "microwaves". My system now has a good balance though and I can scan pretty fast with little microwaves.
2. A reference scan with a very low intersection angle (red number) causes a large distortion on the repetition. I thought maybe David uses the intersection angle to help figure out where the rotation axis of the laser is at? A low angle would make estimating the motion less accurate and perhaps warp the results? If this happens dramatically with a very low intersection angle maybe it is happening only slightly with a better intersection angle. Maybe it causes the little problems the keep many scans from fitting?

3. Slight distortion on the repetition scan after it leaves the reference scan area even though remember plane poses is off. Seems the scan has a sweet spot in the center of the reference corner. Maybe from inaccurices in the camera's calibration or is it the intersection angle or is it something mechanical? I think I read that the tsai camera model calibration data is most accurate within a certain space near the calibration markers (our reference corner) So if we scan outside this area which is easily done with planeless scanning we might be amplifying some camera distoritions?

Okay I am not a scientist and I never studied machine vision or robotics and I may be dead wrong on all these ideas and I probalby am stupid. :oops: I dont mind someone telling me im wrong Even if im really way off Id like to know where I am wrong thanks.

Greg
DealExtreme red line laser, Microsoft Lifecam HD-5000
User avatar
Bongobat
Moderator
 
Posts: 1002
Joined: Mon Apr 02, 2007 10:02 am
Location: Ohio, USA

Postby hal » Fri Jan 15, 2010 12:58 am

Greg wrote:The faster the scanning speeds the larger the "microwaves"

Me too, same problem. I need to scan at very low speed to avoid microwaves.

Greg wrote:A reference scan with a very low intersection angle (red number) causes a large distortion on the repetition.

Never had a very little intersection angle, so never experience about this problem. I'll try to reproduce this situation, in the next days. We will see.

Greg wrote:Slight distortion on the repetition scan after it leaves the reference scan area even though remember plane poses is off.

Same for me. Maybe you are right, but I've not an answer. I don't know why this happen.

I want to add: Have you, sometimes, a lag into your scans, during repetition mode? i mean that sometimes, David stop to calculate the 3D cloud and leave some scanned lines empty (if you imagine to scan the empty 90° corner, the problem happen usually near the low-center of the scene and 3/5 lines of pixels are missed).

Best,
Mattia
User avatar
hal
Moderator
 
Posts: 1527
Joined: Thu Mar 01, 2007 4:08 pm
Location: Cavaria con Premezzo (VA), Italy.

PreviousNext

Return to Laser and Camera Hardware

Who is online

Users browsing this forum: No registered users and 0 guests

cron