A Human Machine Interaction Control System for Bionic Robotic Fish Based on Gesture Recognition

Bionic robotic fish is being applied in the field of education. Human machine interaction on bionic robotic fish has attracted more and more attentions. In this paper, we design a human machine interaction system based on gesture recognition for swimming control of bionic robotic fish. The MediaPipe gesture recognition algorithm here was employed through six gesture modes. Each gesture was designed to correspond to a swimming pattern of bionic robotic fish. The gesture recognition algorithm solved the problem of mirroring when implementing left-handed and right-handed motion recognition. It also simplify the process of bionic robotic fish swimming control. Experimental results verify the effectiveness of the human machine interaction method based on gesture recognition for bionic robotic fish.