This has got to be one of the coolest 3D applications I have ever seen, I’m not sure how to describe it exactly but the video below makes it pretty clear what this program is capable of. Insane!
“Takeo Igarashi from the University of Tokyo has a very impressive java applet/program, called Teddy, which he describes as ‘A Sketching Interface for 3D Freeform Design’, and basically allows you to sketch in simple 2D and have it automatically converted to full 3D. The tool is certainly very impressive and there is a demonstration video available. The end product looks like a hand-drawn object instead of the usual clinical, perfect 3D objects that are designed using standard rendering tools.”
Just saw this over at slashdot and felt it had to go up here – as you’ve probably guessed, I’m a big google fan and I think it’s great to see these types of software moving into the public domain. Given the amount of work and research that has gone into character recognition there is no reason that cheap solutions for document scanning shouldn’t have been available for a long time. I can’t wait for more pattern recognition code to be released into the wild.
“Google recently released Tesseract as open source. Originally developed at the HP Labs from 1985-1995, it has been touted as one of the most accurate Optical Character Recognition (OCR) programs available. Having sat on the shelf gathering dust for so many years, Google cleaned up some of the more outdated portions of the code and released it for general consumption. You can download Tesseract over at Sourceforge.
I’ve got to try this out sometime, and seeing as it fits nicely into my work with 3D face recognition I don’t even have to do it on my own time 
Create your own super hi-tech 3-D laser scanner. Using just a laser pointer, wine glass, rotating platform, and a digital video camera, you can make accurate 3-D models of a object or person.
- Convert the video to an avi.
- Use an edge detection algorithm to find the location of the laser line.
- Reconstruct your 3-D model
Update: I’ve had trouble downloading the m-file so when I got it down I put it up here for anyone to grab.
Thanks to David for the link to this cool demonstration of the face tracking abilities of logitechs quickcam.
I saw an article a while back that detailed a cop cruiser in LA using a new Licence Plate Recognition system which was capable of reading in 30-60 plates a second and sending them through the cars data system to check for outstanding warrants, stolen vehicles, parking tickets etc. Sounded like a great use of technology to imporove law enforcement – however this story in Wired magazine gives a different perspective on the technology, in particular what happens when the price drops to a level where your local supermarket chain can install these things in their carparks?
In recent years, police around the country have started to use powerful infrared cameras to read plates and catch carjackers and ticket scofflaws. But the technology will soon migrate into the private sector, and morph into a tool for tracking individual motorists’ movements, says former policeman Andy Bucholz, who’s on the board of Virginia-based G2 Tactics, a manufacturer of the technology.
Giant data-tracking firms such as ChoicePoint, Accurint and Acxiom already collect detailed personal and financial information on millions of Americans. Once they discover how lucrative it is to know where a person goes between the supermarket, for example, and the strip club, the LPR industry could explode, says Bucholz.
Privacy advocates worry that Bucholz, who wants to sell LPR data to consumer data brokers like ChoicePoint, knows what he’s talking about.
Complexity is everywhere. It’s a structural and organizational
principle that reaches almost every field imaginable, from genetics and
social networks to food webs and stock markets. Contemporary scientific
and technological accomplishments—including mapping the human
genome, decoding neural networks and opening up the ocean to
exploration—have seen our ability to generate and acquire
information outpace our ability to make sense of it. With a surfeit of
facts and few ways to synthesize them, “meaningful information” quickly
becomes an oxymoron.
I really was just interested in the pretty pictures part though They’ve got images of neural networks from a mouse brain, the interconnecting structure of trees/leaves/branches and my favourite the “The World of Music: SDP Layout of High Dimensional Data” – this comes from Yahoo! music which I’m actually a paying member of… so I guess I helped to make some of these little interconnecting lines
More from the robotics arena, researchers from Norway have developed a robotic snake which is 3 m long and weighs 70 kg. The robot is propelled by 20 hydraulic motors which are powered directly from pressurized water within a firehose – apparently there is enough energy available to lift a car or break through a wall.
“It is much like the grab on an excavator where different joints and movements are co-ordinated by the operator. In this instance, the operator is the computer,” says Pål Liljebäck of SINTEF. “There are angle sensors in each joint, and we can decide with conplete accuracy the angle that we want in the joints. A camera in the snake’s head makes operating the snake like driving a remote-controlled car. The operator can tell the snake to move from A to B, and the snake works out on its own how to accomplish this. It knows how to cross a pile of materials, climb down on the back side and twist itself round objects in order to get footing.”
A really interesting concept, these researchers over at Columbia University are working on image processing algorithms that harness the power of the human visual processing to perform activity detection.
The brain emits a signal as soon as it sees something interesting, and that “aha” signal can be detected by an electroencephalogram, or EEG cap. While users sift through streaming images or video footage, the technology tags the images that elicit a signal, and ranks them in order of the strength of the neural signatures. Afterwards, the user can examine only the information that their brains identified as important, instead of wading through thousands of images.
This research was funded by DARPA in order to help operatives sort through reams of image data quickly, however I can see the results being extremely useful in identifying the areas in which current computer vision systems are currently lacking.
