In recent years,with the development of artificial intellgence technology,autonomous intelligent robots that can independently complete tasks have become a research hotspot.Visual perception and gait control,as important means for robots to complete various tasks,are technology points that need to be continuously deepened and improved. In order to expand the autonomous intelligence of small humanoid robots,this paper explores the applications of visual perception and gait control in tasks that simulate human life scenarios(i.e.,the U-shape competition)from the perspective of the International Competition of Autonomous Running Robots.Specifically,this paper first provides a brief introduction to the competition environment and software and hardware platforms,and then analyzes the applications and commonly used classical algorithms of visual perception and gait control in the competition.Next,this paper takes the up and down pole as an example,introduces the algorithms used by the participating teams in the U- shape competition,and conducts analysis and comparison.This paper also shows the excellent achievements of the participating teams over the years,analyzes the task success of the competition,and points out the research challenges and possible solutions.Finally,this paper summarizes the development direction of future competitions.