Information Express of Soft Bionic Finger Based on Visual Recognition

摘要

Soft robots have developed rapidly because of their own structure, adaptability and safety, which can be applied in different fields. However, the human robot interaction of soft robots faces enormous challenges. In this paper, we take the soft bionic finger (SBF) as the research object, design a multi chamber SBF, and propose a visual recognition to drive SBF. The bending information of the finger is converted into the information of driving air pressure, and drive the SBF. The SBF can simulate the bending of human finger and replace human finger to express some information. Finally, in the visual bending experiment, it is verified that the error between the bending angle of the human finger and the bending angle of the SBF is within a certain allowable range. It shows the feasibility of driving SBF to express information based on visual recognition, and also reflects strong human robot interaction.