Detection of outliers in crowdsourced GPS traces

Nowadays, crowdsourced GPS data are widely available in a huge amount. A number of people recording them has been increasing gradually, especially during sport and spare time activities. The traces are made openly available and popularized on social networks, blogs, sport and touristic associations’ websites. However, their current use is limited to very basic metric analysis like total time of a trace, average speed, average elevation, etc. The main reasons for that are a high variation of spatial quality from a point to a point composing a trace and a need for referential data for evaluation of their quality. In this paper we present a novel approach for filtering and detection of outliers in crowdsourced GPS traces in order to assess their spatial quality intrinsically and make them more suitable for more advanced uses such as updating referential road network of French Mapping Agency – IGN. In addition, we propose a new definition of an outlier in GPS data, adapted to intrinsic assessment of spatial quality.