This paper proposes a method for topologically visualizing the perceptual information of a partner robot. First, we explain the methods for human detection, human motion extraction, and object recognition. Next, we explain the perceptual system of the robot based on the detected human and objects. We propose a topological visualization method based on a spring-mass-damper system according to the perceptual information. Finally, we show several experimental results of the proposed method, and the proposed method enables a human to understand what the robot perceives in the interaction with the human and environment.