Sound Source Localization and Interaction based Human Searching Robot under Disaster Environment
Bin Zhang, Masahide Kaneko, Hun‐ok Lim · 2019
In order to find out a human being under disaster environment where visual information is unavailable, this paper proposes a human searching method based on sound source localization and interaction. Especially the interactions between human and robot are designed efficiently to improve the detection accuracy. Let the robot call out the human beings and detect the change of the auditory information in the environment before and after the human gives a response. As important information, the responding time is especially utilized. The human can be located by analyzing the changed auditory information. Through different ways to interact with the human, this method can deal with complex disaster environment, e.g. asking the human to make a sound at a particular time. In this way, the human can be finally searched out even if human recognition methods cannot be used, or the sound is not a human voice.