Modeling of Car Driver Cognitive Process Based on Hazard Perception in Front of Car

Intention estimation is a way to realize an effective driver support. The construction of driving behavior model is an important step for its realization through a model based prediction of driver’s behavior in real driving situation. For the purpose, we constructed a computational model of hazard perception while driving assuming a hazard as a possibility of contact with other objects. We formalized two types of hazard, an explicit hazard and an implicit hazard based on the position probability distribution of object. We compared a reconstructed eye movement by the model and a human eye movement.