Browse Prior Art Database

Leverage visual clues to derive and maintain GPS position

IP.com Disclosure Number: IPCOM000202306D
Publication Date: 2010-Dec-13
Document File: 2 page(s) / 35K

Publishing Venue

The IP.com Prior Art Database

Abstract

The disclosed invention is a method to determine a user’s current location by comparing captured images with images that are already available through web-based repositories in relation to known locations.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

Leverage visual clues to derive and maintain GPS position

When traveling with a Global Positioning System (GPS) or other navigational aid that requires satellite signals, a valid signal is not always obtainable due to physical obstructions or weather conditions. However, a vast collection of visual data is always available. Current image recognition software is sophisticated enough that a system could use it to pinpoint a user's location without a satellite signal; it could at least assist in cases of intermittent coverage.

For example, street-view images from web-based repositories can be used to determine the location. The order in which the images are captured, along with the timing, can be used to determine not only location but also direction and speed. This information can be used to assist approximating the GPS location when satellite signals are permanently or temporarily unavailable. In the event that the satellite signal is intermittent, the visual clues can be used to help fill in the gaps and provide a more seamless navigation to a destination without losing the signal or the missing directions.

The disclosed invention is a method to determine a user's current location by comparing captured images with images that are already available in relation to known locations. One component of the invention is a camera positioned in or on a vehicle or person that has an interface represented on the GPS device. The camera captures images and compares them to the images (e.g., street view) readily available from web-based image repositories to find reasonable matches. A successive set of images increases the accuracy of the match, which, in turn, confirms the loc...