Stealth tool for Street View images
Ein neues Tool soll ab sofort Personen, die für Street-View gefilmt werden, unsichtbar machen und den Hintergrund automatisch ersetzen. Dies das ehrgeizige Ziel von Forschern der University of California in San Diego.
A team led by Serge Belognie and Arturo Flores has just presented a corresponding system. It recognizes people who appear in photos of a street scene, deletes them and automatically replaces the background, as Pressetext.ch writes on Friday. The Street View service from Google Maps in particular could benefit from this technology.
Google has been taken to court several times in the past for data protection reasons. The clothing, shape and size of a person's body are still recognizable. Recognition in connection with the address at which the person was photographed is therefore effortless in most cases. This is where the Californian scientists' system could help.
However, a complete removal of all people from Street View is technically very unlikely. This is because a computer first has to learn what people are, as Bastian Leibe from the Rheinisch-Westfälische Technische Hochschule Aachen told Pressetext. Some other problems will probably still have to be solved, such as the creation of bizarre scenes by retouching people. For example, stray dogs on a leash or bicycles without riders.
www.pressetext.ch
Google has been taken to court several times in the past for data protection reasons. The clothing, shape and size of a person's body are still recognizable. Recognition in connection with the address at which the person was photographed is therefore effortless in most cases. This is where the Californian scientists' system could help.
However, a complete removal of all people from Street View is technically very unlikely. This is because a computer first has to learn what people are, as Bastian Leibe from the Rheinisch-Westfälische Technische Hochschule Aachen told Pressetext. Some other problems will probably still have to be solved, such as the creation of bizarre scenes by retouching people. For example, stray dogs on a leash or bicycles without riders.
www.pressetext.ch
