VAST Challenge 2017: Mini-challenge 1

Abstract

In this paper, we addressed the mini-challenge I. To efficiently identify odd behaviors and general patterns, visual analytics was used to parse trajectory data of vehicles and support spatial analyses. Firstly, the adjacent relations among spatial objects were extracted and then simplified as a topological graph. Base on the topological representation, cluster methods and spatio-temporal visualization were utilized to conduct a comprehensive analysis. Through visual analytics on the topological graph, we found general and regular behavior patterns of passages in the preserve and recognize outliers that reflected odd behaviors.