Real-time crowd formation control in virtual scenes

Wu He, Gaofeng Pan*, Xuan Wang, Jim X. Chen

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)

Abstract

This paper presents a formation control system that can be applied in many areas under defined conditions. The system allows users to simplify interaction and quickly complete formation arrangements in virtual scenarios. At the same time, the system allows the user to input specific formations to meet requirements, such as war formations in video games or formation arrangements for stage performances. It also recognizes user input through a human–computer interaction method based on gesture recognition and finds the best corresponding formation from a library of predefined and user-defined formations. The system ensures real-time user interaction and control by automatically scaling the group size and controlling the group movement according to the unit density. During formation changes, the minimum energy principle is applied to determine the ideal position of each agent in the target formation to increase the efficiency of its movement. In addition, formation stability is maintained during collision avoidance by combining potential obstacle fields with a rigid formation network. Finally, the proposed formation system is tested qualitatively and compared quantitatively with the execution efficiency of some existing interactive formation methods. The experimental results show that the method can effectively implement formation control in a virtual scene. It can automatically avoid obstacles and maintain the formation during the formation morphing process.

Original languageEnglish
Article number102662
JournalSimulation Modelling Practice and Theory
Volume122
DOIs
Publication statusPublished - Jan 2023

Keywords

  • Collision avoidance
  • Crowd simulation
  • Formation transform
  • Human–computer interaction

Fingerprint

Dive into the research topics of 'Real-time crowd formation control in virtual scenes'. Together they form a unique fingerprint.

Cite this