Extracting scenic routes from VGI data sources

Abstract

This paper presents a method to extract traversed scenic routes from Volunteered Geographic Information (VGI) data sources. We used the footprints of images posted on Panoramio and Flickr, as well as websites where users uploaded tracks of traversed scenic routes. The state of California has been chosen as the test region. We discuss pros and cons of these VGI data sources for this task and describe the necessary steps to extract scenic routes. Measures for attributes of the route and the surrounding scenery can then be obtained for scenic routes and their corresponding fastest routes. This information will be utilized to identify relevant attributes that characterize scenicness. These attributes could in consequence be used as optimization criteria to search for scenic routes in trip planning applications.