Humanoid robot runs maze mode using depth-first traversal algorithm

This paper focuses on the humanoid robot walking in the maze. In this research, we proposed the depth-first traversal algorithm for the maze searching with the single-view model and sonar obstacle avoidance theory then follow the “turn right first” principle to successfully avoid obstacles and efficiently walk out of the maze. The superiority of the proposed algorithm is that it can be for various complex mazes. In the three-dimensional maze, the visual system of the NAO robot was firstly used to perceive its surrounding environment, and then the image processing technology was used to identify the position of the surrounding obstacles. After that, the NAO robot can successfully avoid obstacles and walk out of the maze. Today, intelligent robots have a wide range of applications. In order to allow them to quickly integrate into our daily lives, they need to be able to recognize obstacles and walk freely like humans. This requires robots equipped with image processing technology which is able to help robots identify obstacles. During walking, robot will comply right turn in the first. We will use sonar to perceive the obstacles on the left and right sides. And at the turn image processing will be used probe obstacles at right and left. Finally, they will preserve memory of what they have been walked. The experimental results indicate that this method provides a reliable guarantee for the NAO robot to successfully avoid obstacles and get out of the maze.