부산대학교 도서관

Dense crowd simulations are being increasingly used for predicting congestion, planning protests and rallies, and threat assessment. It has been observed that human behavior is uncontrollable and often unpredictable; however, human motion is quite flexible over obstacles in their path and turn as per the need. These motions or the paths they take are majorly based on their individual personal space. Upon introducing these entities to a heterogeneous environment with vehicles, bikes, shopping carts and strollers, each with their own path and their own personal space, human motion drastically shift. Here we introduce a new set of models intended to study such behavior and try to mimic such situations.