ABSTRACT

ABSTRACT: We are mostly concerned with path finding algorithms and decision making in the crowd. We present crowd simulation approach for virtual museum. Visitors in the museum are simulated as agents that are influenced by attractors and markers. Markers are used for collision avoidance and for attractors we use potential fields. Potential fields are widely used in robotics and real time games. They are important for crowd simulations and agent based systems. We present path planning algorithm that is based on decision making extracted from the observation of the real crowds. Our algorithm uses traditional potential fields approach with additional directional vectors. Agents decision process is based on personality and behavior of a real crowd. Agents need to choose next step based on their goal, intentions, potential field force in the neighborhood and collision avoidance strategy. Our approach combines well known techniques, but variables are added to create a realistic result.