Integration with Text Lets Us Get Value Added Images

Like the new Geographic Web layer in Google Earth adds a new value to geographic maps, the FotoTagger technology turns potentially every image in the Web to the source of useful knowledge.

Today I finally tried the new Geographic Web layer in Google Earth announced by Google a few days ago. If you still are unaware about this feature, here it is.

You fly to a destination, click an icon representing a Wikipedia article, Panoramio photo, or Google Earth Community, and see articles, pictures or posts about this place. I think it's a great thing because being combined with different media geographic maps gain a new value.

Now you can use them not just for finding a restaurant or your own house, but use it for self-education. There are still a lot of places without any reference to Wikipedia, but with today's pace of collaborative content creation, I believe this goal can be achieved in the foreseeable future.

Adding a new layer, I would say, a semantic layer, to a geographic map is very close to the FotoTagger ideas.

However, we, in Cogitum, go far beyond pure geographic application and see an opportunity to add the semantic layer to virtually any picture and turn it from a nice looking image into a source of useful knowledge.

Let me give you an example. This is how a usual photo can look.

Now switch the semantic layer on to see the difference. (You can find much more photos on the FotoTagger Galleries that contains many pictures in various fields submitted by users.)

The one and the same picture being combined with a text gains a new value and in some cases, even a new meaning. Look at what's happening with images in the Web. Tagging of visual content has already become very popular.

When Google was releasing the Google Image Labeler game, Luis von Ahn, the father of the game said all images on Google Images could be labeled in two months. So it's just the matter of time when all images in the Web will be tagged with object-specific tags. By the way, personally, tagging images inside is much more fun than in a Google-like manner. Believe me or try by yourself.

Posted on December 15, 2006 by Alex Masycheff

See all posts