Want Your Shapeshot?

Strolling around Rapid 2012 we encountered an unusual service named ShapeShot. What do they do? Well, they shoot your shape. 

 

ShapeShot is a project from Baltimore-based Direct Dimensions, a consultancy that specializes in 3D scanning projects. That's what ShapeShot does: it scans your face and places it into an online repository. They feel it's like "next generation portrait photography". 

 

The ShapeShot machinery is composed of hardware and software. The hardware is a flash-lit four-camera setup in which your head image is captured. The software is powerful 3D analysis programs that convert the four images of your head into a surprisingly accurate 3D model. 

 

The model is then placed in their repository and you'll get a link to go online and view it. But that's not all. They provide the ability to download the model in digital form (at this time OBJ format only, but it sounds like they'll be providing additional formats soon.) They're also considering adding value features such as preparing printable 3D models from the face, thickening walls, or even hooking up to existing print services such as Cubify, i.Materialise, Shapeways, etc. 

 

The website now has 4500+ faces in it, some of which are marked available to the public and some remain private. 

 

While this is all quite fascinating, we wondered about the purpose of this repository. What would people do with it? How would they monetize the service? They said it is merely an experiment at this point and they're seeking partners to develop applications on top of their face repository. 

 

Any takers? 

 

Via ShapeShot

Be An Action Hero!

There are quite a few specialized 3D print services emerging in the consumer space, but we kinda like this one: Firebox. Why? Because you can easily (well, by paying their fee of £79.99, or USD$127) you can get your own head on a real action figure. 

 

The process is simple: take two pictures of your head (or someone else's, we suppose) and upload them to Firebox. They'll take the images and develop a 3D model of your head. They'll 3D print it and ship it to you along with the real action figure, which can be one of Wonder Woman, Superman, The Joker, Batman or Batgirl. All you need to do is surgically remove the original superhero's head and replace it with your own. The process takes three weeks to produce and ship the head and figure, and about 30 seconds to replace the head. 

 

Wait, you say, isn't it copyright infringement to use those characters? We think not because Firebox is simply shipping commercially obtained action figures. Aren't they? 

 

Via Firebox

Tutorial: Fix Your Messed Up 3D Scan

If you ever try to do 3D scanning at home using the rudimentary tools now available, you'll quickly realize it isn't particularly easy. Sure, the capture part of the job is relatively straightforward, but the clean up phase is often quite difficult. 

 

Typical raw scans contain not only stray data points and somewhat inaccurately shaped surfaces, but also the model usually is full of holes. Holes are not good when you're trying to 3D print a model, since the printing software will be confused. Repairs are necessary. 

 

The repair process can be very tedious and often requires use of obscure features in several software tools. How can you learn the right way to repair a messed-up scan? A great place to start would be to watch Tony Buser's 23 minute video tutorial, which takes you through the entire repair process. You'll see screen capture and narration explaining precisely how to get the job done. Tony makes use of several free or open source 3D software utilities that you can easily use at home. 

 

Via Vimeo

3Dify From Matherix Labs

Some weeks ago we wrote on a breakthrough in inexpensive 3D scanning by ReconstructMe, which permits true, 360-degree 3D capture of 3D models of real objects. While ReconstructMe works quite well, it requires a super-powerful graphics card in your Windows PC simply because it's producing the 3D model in real time as you scan. 

 

Realtime scanning is highly desirable as it allows the scanning operator to ensure they capture every hidden niche of the scanned object during the scan. But it does put up a small barrier to those who don't have the required equipment. 

 

Enter 3Dify from Matherix Labs. 

 

This (also Windows) tool again uses the Kinect to capture 3D information, but in a slightly different way. It's not real time - you record a 3D movie of your subject by walking around while pointing the Kinect at the (hopefully) motionless subject. This recorded video is then analyzed post-capture by 3Dify and eventually transformed into a full 3D model. This model can then be tweaked and exported for use in other applications - such as 3D printers. 

 

This approach gets around the hardware requirements, but it's possible that you may have to repeat your 3D scan a few times to ensure you have captured the entire shape because you won't know if it worked until 3Dify completes its processing. 

 

Via Matherix Labs and Ponoko

Human Cloning in Akihabara

We've written about various methods of scanning yourself in 3D, then replicating your face, hands or favorite body part on a 3D printer. But what if you didn't need a 3D printer and could simply walk into a retail location and get it done on the spot? That's precisely what's going on in Tokyo's Akihabara at The Clone Factory. 

 

Danny Choo tries out this service, which apparently costs the nearly outrageous sum of 138,000 yen, which is approximately USD$1800. For that you'll get a "sit down" scan of your head, which is then placed on top of a figurine. The resulting facial detail is a little scary, as you'll see in the numerous images in the report. 

 

The process seems pretty straightforward: after your head is 3D scanned from a comfy seat, the Clone Factory operator prepares a 3D model using Autodesk Maya. Choo says, "They need a while to prepare my data", as is almost always the case on real-life scans. Finally the model is printed on a ZCorp Zprinter 650. While the 650 is capable of printing in color, we're thinking there may have been some post-print hand painting to achieve the amazing lifelike finish. Nevertheless, it's a great result!

 

But here's the question: would you partake of this service at that price?

 

Via Culture Japan/Danny Choo (Hat tip to John)

3D Printing Video Controversy

A 3D printing video from National Geographic went viral on the interwebs last week and quickly generated some controversy among 3D printing enthusiasts. 

 

In the video, physicist David Kaplan visited ZCorp HQ to "find out whether they can print a crescent wrench". Kaplan was shown a selection of amazing items printed on ZCorp devices, including multicolored objects with moving parts, as well as the standard explanation of how 3D printing actually works. 

 

Then the interesting part occurred. Kaplan's substantial wrench was put under a hand scanner, capable of capturing highly detailed 3D information. It then appeared as if this captured data was then converted into a proper 3D model, "creating an image that will be sent to the printer". 

 

Then, just like you'd see on a TV cooking show, the printer was opened and a printed wrench was extracted from the powder. While the wrench was demonstrated to be strong enough for actual use, questions immediately arose. 

 

We're wondering how a hand scanner can discern moving parts, such as are found in a crescent wrench? We know of no technology that could do such a thing, particularly when the video showed no part motion during the scan. Viewers were left with the impression that you could simply wave your scanner at anything and hope to replicate it immediately - including hidden internal moving parts, and by extension even electronics! 

 

That's just not possible. This was noticed and challenged by several sources, including on Jon Udell's blog, where a discussion erupted (link below). 

 

Scan captures are usually nowhere near printable state, and always require at least minor if not major editing and conversion before printing can be attempted. Worse, a given object's geometry might not be printable on a given 3D printer - the right 3D print technology must match the object. 

 

Providing general awareness of 3D printing is a good thing, but leading people to unrealistic expectations is not helpful. We fear that a growing abundance of such memes will eventually cause a backlash towards 3D printing, as folks will soon discover that you can't simply replicate just anything. 

 

Correction: you can't replicate just anything - yet. 

  

Via YouTube and Jon Udell

Photofly Makes 3D Models Easy

There are several ways to produce 3D models of real-life objects and Autodesk has provided another one: Photofly. This is a service and software combination, in which you take around 50 overlapping photographs of your subject and then upload them to Photofly. The service then grinds through some very serious processing to produce detailed 3D models. Models you can 3D print, with a bit of fixing up (like adding a flat base, etc.)

 

You can not only capture 3D models of objects, but also interior rooms, whole buildings and of course, heads and faces. There are a number of constraints on the imagery, including correct exposure, avoiding unnecessary backgrounds, movement, obscuring objects, transparencies and shinies, etc. However, this still leaves a great many possibilities for scanning. 

 

It works through a Windows-only program, the Photo Scene Editor, now available for download. It's only a preview and will expire at the end of 2012, however, but that allows plenty of time to produce a lot of models. At the moment they can produce point clouds, but apparently will soon release a version that produces a true 3D mesh. 

 

Via ExtremeTech, YouTube and Autodesk

3D Scanning by UAV

You may have guessed we're fascinated with new ways of capturing 3D models for printing, and this week we've found a rather unusual method: UAV images. Researchers at Ecole Polytechnique Fédérale de Lausanne's Computer Vision Laboratory have developed software to combine thousands of images gathered by UAVs into very usable 3D models. 

 

The software, produced by spin-off company Pix4D, enables the creation of very detailed, large-scale 3D map models at very low cost. Images are loaded up to a cloud-based service that interprets them using vision algorithms that mimic human vision processing. Within 30 minutes you will have a completed 3D model in your hands. 

 

Pix4D suggests several applications for this new tech, including the ability for farmers to quickly inspect their crop area by tossing a UAV into the air to fly over their fields and collect images. 

 

We think this could be very useful for producing 3D printable models, too. Look up - that UAV might be capturing your neighborhood! 

 

Via KurzweilAI