Attention Guidance for Autonomous Vehicles Based on Type-2 Fuzzy Sets

Visual perception is an important part in the system of autonomous vehicles. The choice of image focus will directly affect decision-making and driving safety, so how to analyze the attention problem and select the appropriate attention area is of great significance. In this paper, an attention guidance for autonomous vehicles is proposed, which can help the system to understand attention areas. According to empirical analysis from different drivers, attention allocations of human drivers are summarized, and membership functions are established. Based on the theory of type-2 fuzzy sets and fuzzy comprehensive evaluation, the rationality of three drivers' visual focus are evaluated in the example.