A cognizing topology map building in a wide-range indoor structural environment

When the exploration scope of the robot is expanded to large regions, the problem becomes more prominent. In this paper, artificial labels are used to provide semantic concepts and relations of the surrounding, which can solve the complexity and limitations of semantic recognition and scene understanding only with robot\´s vision. Imitating spatial cognizing mechanism of human, the robot constantly receives the information of artificial labels at cognitive-guide points in a wide range of structured environment to achieve the perception of the environment and robot navigation. Immune network algorithm is used to form the environmental awareness mechanism with "distributed representation". Color recognition and SIFT feature matching algorithm are fused to achieve the memory and cognition of scenario tag. Then the cognition-guide-action based cognizing topology map is built. Along with the continuously abundant map, the robot need not rely on the artificial label forever, and it can plan path and navigate freely. Experiment results show that the artificial label designed in this paper can improve the cognitive ability of the robot, navigate the robot in the case of indoor structural environment, and build the cognizing topology map favorably.