An example is the David-3D setup. You can use 'normal' projectors - but obviously pico projectors are cheap (relative to the past anyway) and highly portable..
You can combine such a setup with a turntable as well ...
What it wont do well - is scene based scanning of an entire room for example. Thats where a Kinect will excel.
Lawsy here on the forums has a David SLS kit, his posted some examples, but not sure how much his played with it. Theres more info on http://www.david-3d.com/ and specifically http://wiki.david-3d.com/david3_user_ma … ured_light . The barrier to entry here is the projector - I would say the full kit including projector is probably very comparable pricewise to other equivalent solutions personally - but you can always 'bring your own projector' and get the software only for about 2/3rds the price of the MBI solution.
Not saying its a silver bullet - or better/worse than something else - but it does appear to provide higher texture capture than a straight laser/video capture system.
]]>Thanks for the great info. The sense seems great and easy to use from what I see, but it may not net the decent resolution that I would want. Really I want to scan camera bodies and accessories. I'd like to have the flexibility of doing larger objects also, and not be trapped in such a small work space as the makerbot. So if I go the kinect route, for windows, you said I have to be closer to an object. That's no problem, but will it still do large objects? I just would have to scan a lot closer right? As far as the auto stitching and software what's the best choice for that? Seems like the sense is dead easy to use, but that may be a double edge sword leaving out advanced features that are probably needed for certain objects. I have the money and ready to buy but want to make a good decision.
Thanks a lot, much appreciated.
DiyE
]]>Not at all. You can go buy a Kinect and use it with your PC. No XBox needed.
One word of caution on this: There are 2 versions of the Kinect, one for Windows, and one for the Xbox 360. The Windows version only works with the Microsoft Kinect SDK drivers, so it limits the number of software that will work with it to just about ReconstructMe (as far as I know).
The Kinect for Xbox 360 works with the SDK drivers AND with the OpenNi drivers which most of the other 3D scanning programs use, so you'll be able to use it with software like Scenect or kScan 3D.
There is some key hardware differences as well - the Windows Kinect is designed for 20cm range work - i.e, having it on top of a monitor and sitting at a desk. It has much greater resolution as a result in any closeup scanning work. The xbox kinect on the other hand, is optimised to work 'across a room' - it has rudimentary close field support but nowhere near as sensitive as the windows kinect as its designed to observe you from 2m's or more....
Hence why many people are quite specific in stating you need the Kinect for Windows vs the Kinext for Xbox, even given the 'apparent' limitations of the windows version. For scene scanning, the xbox kinect does fine.. for detailed model capture, the kinect for windows provides superior resolution.
]]>The Sense 3D scanner software produces STL, OBJ and PLY files from the final scan. The latter two support per-vertex color output.
This is the same as just about every other 3D scanning software out there.
For printing purposes, the STL file should be all you need, but the reality is that you will more than likely need to import it into some other program to clean it up before you can have a print ready model.
]]>do you have to have an x-box to use the kinect? I've never seen one in action but it looks like that solution would be much cheaper than the sense.
Not at all. You can go buy a Kinect and use it with your PC. No XBox needed.
One word of caution on this: There are 2 versions of the Kinect, one for Windows, and one for the Xbox 360. The Windows version only works with the Microsoft Kinect SDK drivers, so it limits the number of software that will work with it to just about ReconstructMe (as far as I know).
The Kinect for Xbox 360 works with the SDK drivers AND with the OpenNi drivers which most of the other 3D scanning programs use, so you'll be able to use it with software like Scenect or kScan 3D.
So go get a Kinect , plug it in, and have at it. It's cheap fun. However, don't expect much quality wise. Here is a post I made in another thread:
After futzing around with Kinect and kScan, ReconstructMe and Scenect, for a whole weekend, I've come to the realization that, for the type of scanning I want to do, I need a turntable/laser based model.
The resulting scans produced with the Kinect are just too coarse and devoid of small details to be of any use to me. Additionally, this type of scanner is better suited for scanning a "live" scene with large objects, such as a room. They rely on the camera to capture color information to provide the missing detail (the texture). The mesh itself leaves a lot to be desired.
So until I can afford a scanner like the Makerbot Digitizer (or even the MatterForm 3Dscanner), or until such time prices fall to a more affordable level, I think I'll put 3D scanning on the backburner.
To answer the OP's question, I think we need to ask a question in return: What type of objects do you want to scan?
If you want to scan relatively small, static objects, then Sense is not for you, something like the Makerbot Digitizer, NextEngine, MatterForm or a similar type scanner is much better suited to the task.
If, one the other hand, you want to scan live scenes, large objects and such, then Scene will do the job just fine, but then so would any of the Kinect/Asus Xtion solutions. If that doesn't work for you, then you can look at much more expensive solutions with sensors costing upwards of $15-20K or more, or complete solutions ranging in the $100K+.
]]>