Clarifying Origin-Destination Flows Using Force-Directed Edge Bundling Layout

Abstract

Origin-destination (OD) flows have time-varying characteristics and spatial heterogeneity. With the increasing amount of urban travel data, it is very challenging to clarify OD flows through traditional visualization methods. A novel approach based on the force-directed edge bundling (FDEB) algorithm is proposed to visualize OD flows and identify the main corridors in a city. First, we reduce the chaos of OD flows through vertex clustering, and then we use a modified FDEB algorithm to clarify the OD flows. Furthermore, to illustrate the validity of our approach from the perspective of traffic-representation ability, we design three evaluation metrics: local feature richness(LFR) and strength of spatial relationship (SSR), which measure the ability to express the spatial heterogeneity of OD flows, and time characteristic richness (TCR), which measures the ability to express time-varying characteristics in OD flows. Experiments are conducted on real-world automatic vehicle identification (AVI) data that are gathered from Xuancheng, China. The results show that our method can well enhance the representation of spatial heterogeneity and uncover the temporal characteristics hidden in OD flows. INDEX TERMS Edge bundling, OD flow data, traffic desire lines, visualization evaluation.

Uncovering Equity and Travelers’ Behavior on the Expressway: A Case Study of Shandong, China

“…The gravity law indicates that the flow between two areas has positive correlations with node strengths and negative correlations with the distances between the two areas. Based on data mining and spatial analysis, the OD flow exhibits various cluster patterns in many fields [27][28][29]. There are two core categories of OD flow clusters [30].…”