Map construction algorithms: an evaluation through hiking data

We study five existing map construction algorithms, designed and tested with urban vehicle data in mind, and apply them to hiking trajectories with different terrain characteristics. Our main goal is to better understand the existing algorithms and to what extent they apply in a wider context. Indeed, our data differs from the one previously used to evaluate map construction algorithm in several aspects: higher GPS error, narrow and winding paths, and trajectories with its own characteristics in terms of speed or direction. We have chosen four different areas of varied geographic features. For each of them we have considered a set of hiking GPS trajectories, each with a total number of nodes between 38,000 and 288,000. For each algorithm we have analyzed the parameters it uses, and adjusted them to each data set. We present an analysis of the generated maps produced by each algorithm on each data set, and a discussion of the most important artifacts detected. We consider that this analysis sheds new light into the current challenges for map construction algorithms, and will be of help for designing new and better methods.Peer ReviewedPostprint (published version