Detection and tracking of groups in crowd

Abstract

We propose a method to detect and track interacting people by employing a framework based on a Social Force Model (SFM). The method embeds plausible human behaviors to predict interactions in a crowd by iteratively minimizing the error between predictions and measurements. We model people approaching a group and restrict the group formation based on the relative velocity of candidate group members. The detected groups are then tracked by linking their interaction centers over time using a buffered graph-based tracker. We show how the proposed framework outperforms existing group localization techniques on three publicly available datasets, with improvements of up to 13% on group detection.

F-Formation Detection: Individuating Free-Standing Conversational Groups in Images

“…In this case, most of the vision-based approaches perform group tracking, i.e. capturing individuals in movement and maintaining their identity across video frames, understanding how they are partitioned in groups [44,24,53,7,48,22].…”

Dynamic Phase and Group Detection in Pedestrian Crowd Data Using Multiplex Visibility Graphs

“…[13,2,3] investigate anomalous movements in crowd scenes using topology [13] and hidden Markov models [3,2]. On the other hand [9,15,17] model pedestrian group formation using techniques such as social force models and optical flow. None of the existing models use a single data structure for both of these key tasks.…”

Crowd analysis using visual and non-visual sensors, a survey

“…To get rid of the aforementioned noise and clutter, some authors have used body parts models for single and multiple people , which resulted in high accuracy [1,26]. Social force models are also exploited for the detection and tracking of interacting groups of people in crowds [27]. The results were variable subject to density of crowd.…”