Vision Algorithm To Automatically Parse Addresses in Google Street View

By Rebecca Boyle Posted 12.19.2011 at 12:07 pm 8 Comments

Extracting Door Numbers Building numbers come in a huge variety of shapes, colors and sizes, making them difficult to model for machine vision. Netzer et al

Extrapolating numbers and letters from digital images is still a tough task, even for the best computer programmers. But it would be handy to extract business names, or graffiti, or an address from pictures that are already stored online. Aiming to make its Street View service even more accurate, Google would like to extract your house number from its own Street View photo cache.

Say what you will about Street View (and Helicopter View and Amazon View and etc.) — beyond the novelty factor, the images are full of potentially useful data. Using street numbers on the side of a house or business could make navigation programs more accurate, and help motorists or pedestrians find the right door by providing a preview on the Internet or a mobile device. But while handwriting algorithms are pretty advanced, software systems are still limited in their ability to extract information from images. Factors like blurring, lighting and other distortions can be a problem.

To improve matters, researchers at Google and Stanford devised a new feature-learning algorithm for a set of street numbers captured from a Street View database. They used 600,000 images from various countries and extracted the house numbers using a basic visual algorithm. Then the team used Amazon’s Mechanical Turk system to verify the arrangement of the numbers. The result was two sets of images: One with house number images as they appeared, and one with house numbers all resized to the same resolution.

Initially, traditional handcrafted visual learning algorithms didn’t work very well to extract the numbers. Next, the Google-Stanford team tried feature learning algorithms, which use various sets of parameters to learn recognition patterns. The new feature learning methods worked much better than the regular visual learning method: One of the algorithms (a K-means-based feature learning system) achieved 90 percent accuracy, compared to 98 percent for a human.

The system still needs improvement, but it could be useful for extracting number data from billions of images, the researchers say. Ultimately, this could make Street View a lot more accurate. Without a house-number-based view, an address in Street View is a default panorama, which might not actually be the address you want. Type in “30 Rockefeller Plaza,” for instance, and the first thing you see is a chocolatier next to the 30 Rock observation deck. You have to click and drag to see the NBC building.

“With the house number-based view angle, the user will be led to the desired address immediately, without any further interaction needed,” the paper authors explain.

[I Programmer via Slashdot]

Previous Article: Chimps are Unnecessary for Most Research and Should Be Used Sparingly, Report Says

Next Article: Marines in Afghanistan Execute the World's First Cargo Resupply with an Unmanned Helicopter

8 Comments

Link to this comment

Starchild_1

12/19/11 at 12:13 pm

I knew there would have to be some starting point to get addresses into google by using the pictures some time or another. It all depended on the programming to be written. And to think, I actually thought it would be super easy to just plug addresses in and bam, you know where someone lives exactly. Guess I was wrong, I underestimated our technology.

" Science without religion is lame, religion without science is blind." Albert Einstein

Link to this comment

GeeWillikers

12/19/11 at 1:03 pm

My house numbers are 3D and if you look at them from different angles you get different numbers. I try to maintain being anonymous. Take that GOOGLE, ha!

Link to this comment

DCypher

12/19/11 at 1:35 pm

And thus goes the effectiveness of the CAPTCA system to stop spam. If Google is successful I can just see the spammers grabbing the software to read CAPTCHA confirmations and thereby reautomate the entire system.

Link to this comment

Delkomatic

12/19/11 at 2:23 pm

lol ugh spam

Link to this comment

Grunt

12/19/11 at 3:48 pm

The Earth is 3.7 billion years old. The Cosmos is 13.7 billion years old. The Earth was made from the elements of the Cosmos. These elements are found everywhere else in the Cosmos. Life could exist elsewhere in the Cosmos. Old and more intelligent life could exist in the Cosmos.

Take a simple man, from a simple family, at a simple village who lives off the land and has basic tools and no written language. How would explain and keep in his history an experience of a outer space encounter? He would know this is an important event and would keep the story and he would use the vocabulary and knowledge of his time, including local religious superstitions. He would most likely add what he sees to his local religion.

I feel most people deny the possibility of past alien visitation, because they know this could undermine what they believe in their own religions. But ask this person, if there could be aliens far away more advance than us in outer space and this same person can sure, it is possible.

I find it funny how humans can selectively close our minds to other possibilities.

Link to this comment

Grunt

12/19/11 at 3:48 pm

I find it funny how humans can selectively close our minds to other possibilities.

Link to this comment

Grunt

12/20/11 at 10:16 am

I use a computer that is used by many other peoples. I did not write the above 2 comments, fyi.
I do not understand their posting.

........................................

This GOOGLE technology is positive for a better search engine and map making and at times is uncomfortable and all our privacy is becoming smaller and smaller.

Link to this comment