23rd June 2009
Google has released a new research paper exploring the possibilities posed by computers that are able to recognise pictures of landmarks.
Presented at the Computer Vision and Pattern Recognition conference yesterday (June 22nd), the paper describes a technology that allows computers to identify images of over 50,000 global landmarks with an accuracy rate of 80 per cent.
The search engine gathered pictures from Picasa and Panoramio, as well as online tour guide websites, before going on to prune these by using image matching and unsupervised clustering techniques and developing an indexing system for quick image recognition.
"While we've gone a long way towards unlocking the information stored in text on the web, there's still much work to be done unlocking the information stored in pixels," commented Jay Yagnik, head of computer vision research at Google.
Microsoft's research team is also involved with the field of computer vision, particularly in the areas of image and video editing, object class recognition, advanced medical image analysis, real-time stereo vision, visual tracking and remote collaboration.
Category: Internet Trends
Other Internet Trends Articles
Ten billionth tweet shows importance of Twitter to bespoke web development
Cassandra continues rise in popularity for web development
Social media growth important to bespoke web design
Sharing buttons could be key feature of bespoke website design
Web developers 'should consider combining web with mobile'
Top News Categories
Archives
- March 2010
- February 2010
- January 2010
- December 2009
- November 2009
- October 2009
- September 2009
- August 2009
- July 2009
- June 2009
- May 2009
- April 2009
- March 2009
- February 2009
- January 2009
- December 2008
Pure Innovation Ltd