Slashdot Mirror
3D Virtual Reconstructions From Microsoft
Lord Satri writes "New around the corner, Microsoft Live Labs' Photosynth, will 'take a large collection of photos of a place or object, analyzes them for similarities, and displays them in a reconstructed 3-Dimensional space.' There's a demonstrational video and a 'smart photos' example page. From the site Very Spatial: 'The word is that Photosynth will be available for free, at least at first, but no word yet on an exact release date.' I must admit, seems like Photosynth may offer interesting features with an clean interface. This tool will directly compete with Stitcher, and to some extent, Google SketchUp. The virtual world reconstruction tools market is getting crowded, and competition is good. Microsoft doesn't yet have software to tie a photo library with Windows Live Local (Google does), but don't be surprised if it comes to life."
If this software is half as good as the famous: "Dear Aunt, let's set so double the killer delete select all.", then we at least should be able to use it to create Escher like visual paradoxes, but if anyone is hoping to seriously convert a few pictures of themselves into 3d models, they may find themselves in a Dali like nightmare.
You can't handle the truth.
This basically looks like Google Earth based on user photos and not satellite photos. I find it quite interesting, but a little too much of a gimmick right now. Pixel zooming into a picture is NOT the same as diving into the scene and looking around like its a virtual world.
http://religiousfreaks.com/Does anyone know of any open source photo stichers? And by the way, what does NASA use to generate those awesome collages that they produce?
This software could revolutionize buying real estate remotely. Imagine, an agent goes in with a cheap digicam and takes a bunch of shots of the house they're selling. They load them into this software which creates a 3D, navigable model of the house, which someone can browse via a browser plugin.
Sure, this has been around for a while with VRML, but it was complicated and costly for an agent to do. From the looks of this software you can use normal photos as a base. Anyone could create 3D tours with this.
What about cars and people that change from picture to picture? It's not possible to match them. Does their reconstruction algorithm try to erase them, or to merge them?
What happens if you throw some Escher drawings at it?
This guy's the limit!
Deckard: Enhance 224 to 176. Enhance, stop. Move in, stop. Pull out, track right, stop. Center in, pull back. Stop. Track 45 right. Stop. Center and stop. Enhance 34 to 36. Pan right and pull back. Stop. Enhance 34 to 46. Pull back. Wait a minute, go right, stop. Enhance 57 to 19. Track 45 left. Stop. Enhance 15 to 23. Give me a hard copy right there.
Weaselmancer
rediculous.
I have to say, this would be pretty neat - take all of the images on Google Images for instance and be able to take a high-def virtual tour of places around the world.
Be sure to check out PlayAnywhere too - another neat tech that's being made over at Microsoft Research.
Looks like panarama software on crack. Lots of legal implications I would think - depending on how the photo's are shared or linked - since it is taking photo's that you may or may not have shot and combining them all together - the question might be "who owns the final composite?".
Looks amazing though - can't wait to see it come out.
www.wildpad.com
Yes, there is this software: http://user.cs.tu-berlin.de/~nowozin/autopano-sift /
It is GPL'ed but the problem is that it is using a patented algorithm (SIFT features) so it is not free to use in commercial applications without paying I guess.
pooyak.com
This is doable, if this product is for real and sells at a good price it will be great. The only software that can do this right now is very expensive camera tracking tools, like :Bijou
Sig removed because it was obnoxious
you are looking for 'Hugin', generally is does a good job.
--Rant--
however it does suffer from the fact that it is a font-end for a series of command line apps which have widely disparate design paradigms. One particularly annoying app is written in C# and thus one must down the whole of mono crap in order to use it (if memory serves it is a 28 meg download for a 780K application). Also due to the same issues there is not really consistent handling of filetypes, size, colour depth, compression, and error returns. A corollary problem is that the developers are very enamored with algorithms and not so with implementation details so when things go well they go very, very well and when things go wrong... well it's a bit of mystery.
--Rant--
However I must be clear despite all that I complain about hugin it sucks much less than realviz or any other closed source app I've tried.
Nothing in the world is more dangerous than sincere ignorance and conscientious stupidity.
Upload Natalie Portman.
Then fly over her 3D body in realtime. Excellent!
8 of 13 people found this answer helpful. Did you?
A formerly despised and hated company actually ends up doing new neat stuff, whilst a new protaganist takes over, formerly loved as an underdog, treats everyone like crap and becomes hated.
REAL VIZ has been doing this stuff for years. They even have a few movies under their belt where their software has been used. http://www.realviz.com/
I use Hugin Panorama Tools. I think it works really well, though I don't have a lot of experience with other stitchers. It can do panorama stitching, and also correct for barrel distortion and such.
Their website shows wikipedia, not MSN Encarta :)
"goatse man in 3D! Best horror flick imaginable..."
That is a truly disturbing mental image for this early in the day - or for any time of day in fact.
I know that I may get trolled down for this, but it looks like Microsoft has actually created a cool piece of software here. Granted, it is NOT true 3D worlds. However, I have taken enough photographs in places that do overlap that I think it will be a fun gimik. There are a couple of things I am wondering about. Will Microsoft be selling this software, or is it bundled with Vista? What type of processor do you need? How long does it take the computer to do the calculating and create these "Virtual 3D worlds"? From the way it sounded, I am supposing that their will be internet connectivity to build bigger worlds based on what other users have taken. What about differences in cameras and color settings, how does the software determine which is true color?
Still, may be worth poking around with.
...but no word yet on an exact release date.
Maybe it's being bundled with Vista. *snigger*
From what little I can make of everything I read, LiveLabs is more of a think tank that is funded by Microsoft. I don't believe they are even under much if any creative control by MS. I would think of this more like a small startup with an idea and an enormous budget... memories of the dotcom era.
So because of this affiliation, MS comes out looking innovative and creative when it's merely a small team of appearently very creative developers who have probably never touched any code of any of MS's major income generators (Office, Windows, etc).
Sometimes the best solution is to stop wasting time looking for an easy solution.
Why does this feel like MS is about to steal some other companies software and everyones digital media they produce? The video sounded like a MS Labs commercial or marketing an excuse for stealing before public out cry. Expect some EULA that says all our digital media belongs to MS but we can buy a license for it.
hugin does photo-stitching pretty well, I find, and is open source.
did you notice that the guys in the video do not show _how_ the 3d environment was made from the photos, they just give a nice presentation of assembled pictures. they probably spent hours on adjusting and calibrating the images. the automatic recognition of the marker points in the images will never be fully automatic - especially not if you consider all the low quality pictures of amateur photographs. probably there will be the user doing endless clicking as in those 3d reconstruction tools like imodeller (see http://www.imodeller.com/ ) and image modeler ( see http://www.realviz.fr/ ). btw. you need a lot of pictures and a lot of well picked (error less than a pixel) feature points to robustly estimate the camera geometries...
Correct me if I'm wrong but doesn't the wisdom of the ages dictate that first you find a need and then deliver a solution?
It depends on your goal. If your sole goal is to make money, they you are correct. As far as general scientific and technological advancement you are REALLY wrong.
"reality has a well-known liberal bias" - Steven Colbert
Microsoft develops tons of experimental projects that are not meant to see the light of day (at least commercially) at Microsoft Research and MS Live Labs. Some of those are not publicly divulged, but are shown internally to other MS Employees. This is nothing new...
As a Slashdot discussion grows longer, the probability of an analogy involving cars approaches one.
Wake me up when it's over.
Faster! Faster! Faster would be better!
Research and Development
Sometimes you do things, not to fulfill a specific task but to explore a concept or an idea, even. Lots of money is spent this way - not just by venture capitalists or companies looking to make a buck, but by research firms who have an honest interest in progressing the sciences - yes, eventually they will make money but in the short term research like this is important just for the sake of knowlege. Imagine for example the defense application. Send a UAV through a remote hostile location with a camera snapping pictures every tenth of a second. Stitch the pictures together into a battlefield scenario that can be imported into a 3D visualizer. Research leads to products.
Right now this is basic research. Some potentially cool applications but nothing yet. Unsurprisingly, it's Microsoft doing the basic research anymore. I remember when other companies funded such things. Apple used to have an entire skunkworks dedicated to basic and advanced research. Sigh. Well, at least we'll be able to see the new and creative appear from the academic computing centers, it'll just run on Vista first.
Anybody else catch that bottle of Microsoft Brian Wash in the video? At least somebody's got a sense of humor of there...
-mrxak
Onions Will Kill You
Microsoft's tradition of little R&D apps predates the existence of Google. Once upon a time, sandbox.research.microsoft.com was just chock full of little goodies. Microsoft turned maybe 2% of them into products, and liscensed the other to third party companies (bit of trivia - iPod scroll wheel developed by Microsoft Research as a volume control for VoIP phones, they didn't use it and liscensed it out).
The big difference is that Google started adding a limitted level of release support to their betas, and it became a very popular program. Microsoft is definitely leaching the idea that these apps should be consumed by the public at large, but Microsoft Research has been and continues to be the single largest best-funded CS-related research group on the planet, and they come up with some truly amazing stuff.
Apple's venerable QuickTime VR Authoring Studio was once the virtual reality application of choice. It would have been nice to have something like the photo editing machine thingy Rick Deckard (Harrison Volvo) used in Blade Runner built right into it. But sadly, Apple let QTVR Authoring Studio lose itself in time, like tears in rain...
Beauty is in the beholder of the eye.
I know you asked about an Open source stitcher, but there's also Autostitch to have a look at. It's Windows-only, but from what I can tell, their demo version has no time limit, and it does an impressive job with braindead simplicity : select pictures, click go.
OK, I won't be surprised.
I also won't be surprised when slashdotters gush and fawn over Google's product, then go ape-shit over Microsoft "tying" a software product to Windows Live Local.
There was a long discussion of pano tools, both free and commercial, over at dpchallenge.com a while back. That link is to the first page where the discussion starts, way down at the bottom. From that thread, it would appear that, while a major PITA to install and learn to use, Hugin produces results that are typically at least as good as most of the major commercial tools and are far better than many of them.
I can see where this would be a big help in investigations, journalistic, scientific, criminal, etc. Reconstructing a 3-D scene would help understand where people and things were when something happened.
Today there are mic's placed in some high crime areas that identify a gunshot and where it happened. Cameras placed at strategic locations would complete the "picture".
A very nifty tool, not very well known: http://hugin.sourceforge.net/
there's no place like ~
"What is that tower called? Just photograph it.
Photosynth could eventually connect you to everything on the Web related to it."
Replace 'tower' with 'picture of naked girl' and you realise the real possibilities.......
The next logical step (as the algorithms improve, hardware gets faster, and demand grows) will be to do the same with video. See http://www.bigfootencounters.com/files/mk_davis_pg f.gif to see a cursory example of how motion picture data can be used to build a persistent environment.
Another poster earlier in the thread speculated that a real estate agent could photo a house to make a virtual tour. Even better, maybe, would be to just carry a high def video camera of some sort through the house, waving it around to get at least a little bit of footage of everything. With that data, an intelligent program could composite a 3D representation with even fewer blackout spots. Combine this with an accelerometer/gyro field that gives a non-software correlation to the video stream, and it's essentially bulletproof.
In the form demonstrated, this is a fantastic heavy duty software solution, but physical tracking data would both make this job easier and improve the quality.
I suspect that in the near future we will see the following technologies made ubiquitous in cameras:
1. GPS
2. Tilt/Compass
3. Accelerometer/motion tracking for video.
Items 1 and 2 would enable any camera to provide very accurate geo-located data. #3 with video gives you tracking where GPS fails plus the super accurate tracking data needed to take this to the next level.
"But Chairboy, you tool, why would the camera companies go to the expense?"
The features listed have become incredibly cheap (both in cost and power consumption) over the past few years. Within a couple years, it'll probably be hard to NOT have them in one of the shared chipsets the camera manufacturers use, and at that point, why fight it?
(And this is probably off-topic, but what the hell)
:)
Anyone know of software that can take 2 (or more) digital pics of, say, a person's face from slightly different angles and then try to make a 3d model of it? I assume it exists, I just have no idea what it would be called.
Tried googling, and am getting a bajillion results for stuff that just isn't related.
Anyway, it would be really quite interesting to see the tiling software like this coupled with the perspective/parallax type of 3d modelling and ultra high-res photos. Obviously the hardware to do this would be... uh, intense, but it would be very cool to just snap a few shots around my place (or put a camera on a tripod, set it to automatically rotate and snap, then raise up, do it again, etc, all the way from floor to ceiling) and poof, instant 3d model of the entire place. Then let me edit the model to seperate objects - coffee cup from desk, desk from floor and wall, etc. Or hell, build a library of shapes for the software and have the computer do it automatically, calling in a human only when the shapes are too complex/confusing. Heck, could even then apply textures and reflection etc. options based on the photos, create and position light sources etc. The geeklet in me is drooling at this
Since I can't tell them apart, I treat all ACs as the same person.
From the abstract I thought at first this was something like Canoma (developed by Kai Krause's MetaCreations, then bought by Adobe, then - dropped?). With it you could make and texturize(!) 3D models from a photograph. Actually it was even working with comics.
I wrote something about using this technology to build massive 3D maps, with photo repositories such as flickr.
what is nailchipper?
I'm rather curious to see how well their approach scales. For example, what if you just dumped all the 1,853 photos of Times Square from Flickr into their interface? Scaling even more, in the future could one use this to aggregate all the photos in a particular city, or even have a Google Earth-like interface aggregating photos from all over the globe and integrating it with satellite data? There's some interesting computational problems with arise in trying to find correspondence between that many visual features.
I'm also like to see if they can deal with pictures taken at different times of day. I'm guessing it's still too difficult to actually adapt a day image to a night image, so it'd probably just end up treating photos taken at different times of day as different scenes.
Their video is also MUCH better. Much more impressive, they show some very cool features Microsoft did not. Still, both videos only show the User Interface. Not the calculation of the dataset. It is however no secret that Microsoft PhotoSynth is basically this with a different UI. Or maybe completely the same. (Notice that the Microsoft name is both present on the PhotoTour homepage and the paper for SigGRAPH).
Do also read the SigGRAPH paper. This is the actually tech part. http://phototour.cs.washington.edu/Photo_Tourism.
Some interesting facts you'll find there:
- They used a 3.4 GHz computer.
- The Notre Dame example took two weeks to compute.
- only 597 of the 2635 images were used, the rest was discarded.
This definitely isn't of much use to "home" users. You'll need a semi-professional photographer to cover the whole location. If there are some bigger holes in the imagery, you'll probably not get a useful result. If the images don't overlap, how is the software supposed to calculate their relative position?Debian GNU/Linux - apt-get into it.
Or take a picture of anyone you really fancy and find out where they live....
Another application for software like this - doing reconstruction work on historical/artistic/cultural sites.
why stop with photos? as technology continues to leap, having a real-time "computer" navigable model of data via cameras would allow intelligence to capture and process data.
where would these cameras be? why would they have to be mechanical, why not biological cameras? insects, or "bugs", that have been genetically designed to transmit their image data over "wifi". and who would have control of the data? net neutrality?
http://www.epoch-net.org/index.php?option=com_cont ent&task=view&id=46&Itemid=88
Microsoft related website showing image of wikipedia?? http://labs.live.com/photosynth/images/page_6.jpg
Amazing...
WTF am I doing replying to an AC at 5 A.M on a Friday night?
Apparently this publicly announced and publicly available project can be considered R&D? Who knew that such applications which have been around for more than a decade in a commercially available and pretty much final form minus the 'social' aspect, would count as research and development = notice lowercase... that's how much I think of this concept.
Did i really miss something of significance here or is this YASNT (Yet Another Social Networking Toy)?
A fool throws a stone into a well and a thousand sages can not remove it.
IANAL ... and I'm really pleased I'm not! Man, who cares? Is this all people think now?
/. is a tech site, not a friggen legal shit-fest.
"Oh cool. Imagine the fun the lawyers will have with this"
Theres been plenty of cases regarding who owns images when a person or object is out in public. Enough already,
I wasn't talking about who owns a likeness (building or otherwise) - if you watch the video - he specifically says that you might choose to zoom in to an area where you weren't able to capture enough detail. The software then automagically looks out in cyberspace to see who else might have captured the data. When it finds more source art, it recompiles it to the same perspective you were trying to view your original picture with. I'm looking at it from the perspective of a photographer.
/. so pooey on you.
So the question isn't whether or not taking a picture of a certain church or building is legal, but rather - if it auto-recomposites the image for you using sources from all over the place - who owns the final composited image?
It's similar to sampling - you can sample sound up to I think 16beats before you are crossing a legal threshold.
Besides - last I checked, "Your Rights Online" was still a topic on
www.wildpad.com
To have a right to do a thing is not at all the same as to be right in doing it
AutoStitch is a research project from the U of BC, but the demo app is advanced enough that you can just drop a dozen images into the thing and it figures out the rest. Much easier than any other tool I've used.