Make3D - From 2D to 3D in an Instant

Automated generation of 3D content is an interesting and noteworthy problem that seems to be getting more and more attention lately. With the boom of custom user-driven 3D content both online (virtual worlds, 3D chats, etc.) and off (video games) there is an increasing demand for better, cheaper, and more personalized models. Automating model creation, or at least an initial base model, would allow content producers to work faster and in higher volume without sacrificing quality.

Along with our previous reviews of PhotoSynth and VideoTrace we will now add Make3D as a promising example of automated 3D.

Make3D relies on a complex algorithm created by researchers at Stanford University. As users upload images to the Make3D site, algorithm increases it's efficacy by 'learning' more about constructing 3D scenes based on the new 2D information it has. From the site:

The algorithm models both image depth cues as well as the relationships between different parts of the image. Other than assuming that the environment is made up of a number of small planes, our model makes no explicit assumptions about the structure of the scene; this enables the algorithm to generalize well to scenes that were not seen before in the training set.

Check out the gallery for some very cool interactive examples of Make3D's output. Though the site has recently suspended automated new user registration, a note on the site indicates that you can still get access by emailing the webmaster.

As a bonus for researchers and ambitious tinkerers the code is available for download as well!

Digital Preservation in China

The Bureau of Surveying and mapping in China has announced that they are nearly finished measuring and digitizing the last surviving portions of the Great Wall. The digital models are being created from detailed photographic mappings of nine provinces in the northern region of the country.

Furthermore, the traditionally Internet-unfriendly Chinese government will make the models available online.

This is a great example of the preservation benefits of 3D digitizing and modeling. Even though the wall is significantly degraded, future generations will be able to view and reproduce this "Wonder of the World" as it stands at this time.

Maybe Google can convince them to allow the inclusion of these models into their recent efforts to catalog history?

Project Launch: 3D Pendant Configurator

We are happy to announce the successful launch of our latest interactive web project, a 3d configurator and sales application for custom made pendants. Commisioned by Austin & Warburton Jewelers of Ann Arbor, MI, the application was launched at TheParentAndChild.com shortly before the busy holiday shopping season.

From the project page:

We designed a set of customized 3d configurators that independently focused on each of the detailed and unique pendants. Users can load one of the three pieces and configure it to their own specifications. Choices include pendant size and several metal types and colors. A wide cariety of gemstones can also be selected and placed in specific locations. Each configuration choice is updated instantly on the 3d model without a page reload.

The application was fully integrated with the client's existing shopping cart system and provided a significant upgrade from their previous set of static 2D images. The previous month of heavy use has shown the application to be a great success as described in the message below from Craig Warburton, owner and manager of Austin & Warburton Jewelry:

"It is difficult to describe just how much the product configurator has meant to our online business. Cart abandonment has dropped by 50%, returns for selecting undersized birthstones has all but disappeared and sales have grown 50% faster than I ever imagined possible. The 3D format allows the user to see just how thick and heavy the item is, removing the "cheap and flimsy" fear so much online jewelry formats invoke. The service we received was world class in scope, depth and timing. The seamless integration into our shopping cart software is amazing. In one keystroke, the user can place as many as 10 items into the cart, and we gained control otherwise only provided by written warnings when you established "rules" for application of stone sizes and locations. I always felt our website was superior to many others in our product field, but the configurator you created is the new engine of sales and unmatched by anything in the marketplace. I will call on N Formation Design again and again. The value provided far exceeded my every expectation."

Googling the Future

For a great example of the benefits of 3d and virtual environments, take a look at the new addition to Google Earth recently reported in an article by a Bay Area ABC affiliate.

Google Earth has become a great tool for interactively viewing existing structures and topology, but the truly novel step comes in integrating future structures with the existing world. Google has done just that by adding models (provided by Caltrans) of the yet-to-be-completed Bay Bridge over the existing virtual San Francisco Bay.

From the article:

Until now, Google Earth had a policy of showing only completed structures, for an accurate picture of the earth. But along with adding 3D images, Google has decided to go into the fourth dimension -- time. The bridge is the future, Ancient Rome is in the past.

"I would like to do more of the past, more of the future, to make Google Earth not just everywhere, but every when. A time travel machine for people to understand their heritage, and their future," said [Google CTO Michael] Jones.

I hope this is not just an idle PR statement. Google Earth over time (Google History?) would be an amazing and unprecedented educational tool. The amount of data that could be incorporated is almost unfathomable. The movement of land masses, industrialization, perhaps even battle re-enactments. If anybody can do it, I believe it will be these guys. Fingers crossed!

Desktop 3D Scanning

The RealView 3D company has announced a new desktop 3D scanner that will be unveiled at the recently opened CES 2009. It's great to see more competition in this market. Hopefully the prices of these systems will begin to come down! NextEngine seems to have the most popular 'low cost' version which still comes in at $2995 for the bare-bones package. Though we don't have one in house, I have seen demos of the NextEngine system and the quality is very good for the price. Unfortunately, there doesn't seem to be any price or quality info currently available for the RealView scanner. It will be interesting to compare the too once more details are released. A recent Wired article is already predicting a loss for RealView...

In the meantime, why not try out a do-it-yourself solution! The DAVID scanner requires only a handheld laser ( < $30 USD), a webcam (< $50 USD), and a license for the DAVID-laserscanner software (trial version - FREE, pro version, ~$275 USD). There are some demos and production videos on the site that are definitely worth checking out. Pretty amazing for the investment! If this sort of project is up your alley, why not add on a DIY 3D printer for a full-blown homemade prduction pipeline!

Though all this solutions will inevitably require some patching and smoothing by a 3D artist for production quality models, the increasing availability and accessibility of these machines will no doubt decrease the workload for developers and allow us to produce high-quality content faster and more cost-effectively.

3D Nature! - Interactive Vizualization Hits the Big Time

Nature, the preeminent journal of scientific research, has boosted the credibility of interactive 3d for scientific visualization by publishing their first 3d interactive figure. The January 1st issue contains 3d figures of molecular clouds in a paper first-authored by Alyssa A. Goodman, astronomy professor at Harvard University.

Nature makes use of '3d PDF' technology which allows 3d applets to be embedded in PDF document pages, much like we currently embed them in traditional web pages. Try out the published models (and others) for yourself at this page on the site of Goodman's Initiative in Innovative Computing group. Note: You'll need Adobe Reader 8.1.2 or better to view and interact with the 3d modules.

Screenshot:


We won't pretend to comprehend exactly what is going on in this figure, but one thing we do know is 3d - and the fidelity of this model is fantastic. There a few awkward interaction elements that could be smoothed out, but nothing due to limitations of the technology. The most exciting about this publication for us is that the backing of Nature is sure to increase the exposure and enthusiasm for interactive 3d technologies. Many fields that already make use of 3d technologies to visualize data in their labs will greatly benefit from this expanded ability to communicate their findings to readers.

Anyone interested in using 3d PDF technology for their own work should contact us for more detail.

3D Arlington - Leading the Way

A recent article in the Washington Post highlights efforts by Arlington County government to bring the region into the flourishing world of interactive 3d. Resident John Feather is volunteering his time to get these advanced technology initiatives off the ground and making sure they get the attention they deserve.

Much of the article is devoted to area presences in Second Life. Feather (avatar name: Theatre Magic) leads reporters and other interested parties on guided tours of some of these spaces, making sure to show the power and flexibility of the environment:

One recent evening, Theatre was shepherding a reporter toward the county's online office space for a tour when he sent a text message suggesting a detour. "Want to go swimming?" he wrote. He then teleported to a lake surrounded by Japanese maples and did a summersault from a cliff with an animated Buddha nearby. It was a long way from the Arlington office, but what better place to ponder the future of government?

We need more people like John! It's great to see people in previously untapped areas of life getting the interest and enthusiasm for interactive 3d that us pros have been shouting about for years. I found the following note most interesting:

In November, he started working on a 3D map of Arlington's major buildings. Touching images on the map calls up Web pages about them, and he and his colleagues want to add real-time rent data and detailed visuals from architects and developers so that "when you click on that building, you go in the door."

This sounds like a great idea that even non-techies can see the benefit of. Incorporating spatial data (buildings and locations) with other info (rent data, etc.) will allow people to explore data in a new and relevant way while helping them to answer important questions (where should I live?).

Keep it up John! We applaud your enthusiasm and hope that it inspires others. Let us know if there is anything we can do to help!

EmergingTech: ILoveSketch

ILoveSketch is a very cool application recently announced by the Dynamic Graphics Project (DGP) lab at the University of Toronto. In short, it makes sketching in 3D as intuitive as I have ever seen it. Anyone who has ever tried 3D modeling, especially an organic object, knows that the process is far from straight-forward. Sketches are usually used only as 2d reference points that have to be viewed in a single viewport.

Though ILoveSketch doesn't translate sketches directly into models, it's easy to see how helpful it could be in the process. And who knows, maybe the DGP will use this great app as a starting point for something bigger? Take a look at a video of the software in use and see for yourself!

From the site - ILoveSketch is:

"A 3D curve sketching system that captures some of the affordances of pen and paper for professional designers, allowing them to iterate directly on concept 3D curve models. The system coherently integrates existing techniques of sketch-based interaction with a number of novel and enhanced features. Novel contributions of the system include automatic view rotation to improve curve sketchability, an axis widget for sketch surface selection, and implicitly inferred changes between sketching techniques. We also improve on a number of existing ideas such as a virtual sketchbook, simplified 2D and 3D view navigation, multi-stroke NURBS curve creation, and a cohesive gesture vocabulary."

The Open Source Solution

I've been getting a lot of questions about what the 'best' or 'easiest to learn' 3d modeling program is. After further discussion, it usually turns out that the person really wants to know what the 'cheapest' or 'easiest to obtain' program is. My answers to these people usually come in the form of some sort of open source solution. I've been really surprised at how few people actually know about these solutions or even the open source concept in general.

From the Open Source Initiative homepage: Open source is a development method for software that harnesses the power of distributed peer review and transparency of process. The promise of open source is better quality, higher reliability, more flexibility, lower cost, and an end to predatory vendor lock-in.

Open source allows users to contribute to the development process by writing, reviewing, and optimizing source code. Non-developer users can also contribute by providing bug reports, reviews, or financial support. Everyone gets the advantage of a freely available product that often has less bugs, is easier to use, and has more of a support community than its commercial rivals.

Following are some of my favorite open source solutions. Give them a try and remember to contribute where you can!

3d Modeling/Animation

Blender - Blender Tutorials

K-3D - K-3D Tutorials

Raster Image Creation/Editing

GIMP - GIMP Windows Installer

Vector Image Creation/Editing

Inkscape

Office Tasks (word processing, accounting, etc.)

OpenOffice

Emerging Tech: VideoTrace

Creating great looking real-time ready models of complex objects is getting easier and easier.

Anton van den Hengel of the Australian Centre for Visual Technologies is leading the development of an amazing new process that allows you to easily create 3d models from video. The process - called 'VideoTrace' - debuted last year at SIGGRAPH and involves interactively tracing the desired object over several frames of video.

Check out a very cool demo video here or here.

Image processing algorithms have existed for years that attempt to automatically detect and generate 3d meshes from a set of images. Examples of successful implementations are found in medicine and geology, where data sets involve known components with little variation. The limiting factor in taking this idea to general use has primarily been the ambiguities (jitter, fade, artifacts, etc.) of an undefined visual source.

VideoTrace compensates for these shortcomings by providing an interface for the user to interactively trace the desired object in the video. These tracings give hints to the algorithms about what should be processed and (theoretically) remove the major ambiguities. A paper describing the technology can be found here.

A note on the ACVT VideoTrace page says that they are currently looking for support for commercial development. I sure hope they get it. Even though models of some small objects would still be easier to generate with a low-end 3d scanner, the applications for this technology on a large scale are huge. Think of a bridge between Google Streetwise and Google Earth - users could quickly trace the video of all the houses on their block and have it uploaded for use in Google Earth.... The possibilities are many!

A Valuable 3D Resource

I thought I'd post a resource for anyone looking to improve their 3D modeling skills. The Ego Lounge over at Ego-Farms has a bunch of really great tutorials and comment threads with helpful tips and discussions. It's also a great place to get feedback on your work or help with a particular problem or technique.

Ego-Farms is run by Eric Maslowski, a great guy and ridiculously talented 3D artist. Take a look at his portfolio if you want something to aspire to!