Abstract:With the continuous progress of robot technology, the autonomous navigation ability of robots in unfamiliar environments has become a hot topic of research. Traditional navigation methods rely on known environmental map information or match pre-set nodes to determine specific location information. However, in scenarios lacking prior map information or satellite signals, these methods struggle to handle complex and dynamically changing environments. To solve this challenge, this paper proposes a novel approach based on motion perception and biomimetic brain-like navigation. By simulating the navigation mechanism of the mammalian brain, combined with a bionic polarization orientation system and node detection method based on feature point descriptor, efficient autonomous localization in an unfamiliar environment is realized. This paper conducts both qualitative and quantitative experiments in various environments such as jungles, rural roads, and urban alleys to verify the feasibility and effectiveness of the proposed method. The results of the quantitative experiments show that the absolute positioning error of the proposed method is reduced by 60% compared with OpenRatSLAM. The results of the qualitative experiments indicate that the proposed method has a smaller positioning error in unfamiliar and complex environments and exhibits stronger environmental adaptability.