Vision-Based Environmental Perception and Navigation of Micro-Intelligent Vehicles

摘要

The adjustment of actual environmental traffic flow experiments is complicated and time-consuming. In this paper, a method based on micro-intelligent vehicles (micro-IV) is proposed to overcome these unfavorable factors. Vision-based environmental perception employed should be qualified for real-time and robust characteristics. An active vision approach based on visual selective attention is carried out to search regions of interest more efficiently for traffic light recognition. Corner detection based on histogram is applied for real-time location in autonomous parking. A method based on hierarchical topology maps is proposed to realize the navigation without GPS equipment. Experimental results show that the perception and navigation approaches work efficiently and effectively and micro-IV is suitable for traffic flow experiments.