Spatial Visibility Clustering Analysis in Urban Environments Based on Pedestrians' Mobility Datasets

摘要

In this paper, we propose a Spatial Visibility Clustering (SVC) method estimating the number of clusters (groups) k, based on 3D visible volumes analysis in urban environments. We extend our previous work and propose fast and exact 3D visible volumes analysis in urban scenes based on an analytic solution. We test and analyze the SVC method by using real records of pedestrians' mobility datasets from the city of Melbourne and by setting control points for efficient monitoring and control using a K-means clustering algorithm. By testing large databases, we also propose time zones division for optimal control points.