Collision detection: review of methods and recent advances in crowd simulation

• Main Authors: Mohd. Suaib, Norhaida, Abdullasim, Nazreen, Zolkifly, Iznora Aini, Bade, Abdullah
• Format: Conference or Workshop Item
• Language: English
• Published: 2014
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://eprints.utm.my/id/eprint/61500/1/NorhaidaMohdSuaib2014_CollisionDetectionReviewofMethods.pdf; http://eprints.utm.my/id/eprint/61500/
• Institution: Universiti Teknologi Malaysia

Crowd simulation is a large complex system that visualizes the behavior of crowd entities' movement and their interactions with the virtual environment. Crowd model is usually integrated into a virtual environment to make the environment alive. In the context of agent-based simulation (as in crowd simulation), it encompasses collision checking between moving agents that are present in the same environment. Hence, it is important to design an efficient and yet effective collision detection in crowd simulation. This is to ensure that it is cost effective toward computational processing usage and still produce a believable behavior. This paper presents a study of collision detection techniques in crowd models, and recent advancement to accelerate the process so that in turn, these efforts could also improve the performance and outcome of crowd model in virtual environment applications.