Research on motion behavior control of the carp robot based on computer vision

Bio-robot is a kind of creature controlled by human beings by applying intervention signals through control technology to regulate biological behavior. The animal can be controlled artificially and become a kind of animal robot by stimulating the brain motor area of the animal. To realize the precise control of the carp robot, it is necessary to monitor its motion state and quantify its motion behavior. Therefore, the article proposes a method of applying computer vision technology to detect the control effect of the carp robot. In this study, the method of combining inter-frame difference with Gaussian mixture background modeling was used to extract the target, the mass cancroid tracking algorithm was used to track the target, and the parallax method was used to obtain the three-dimensional motion trajectory and various motion parameters. The carp robots were divided into the Experimental group and the Control group, with 30 fish in each group. In the Experimental group, the computer vision technology was used for real-time detection, and electrical stimulation parameters was adjusted according to the measured data to control the movement of the carp robot. In the Control group, the underwater control experiment was carried out without computer vision technology. The results showed that the success rate of control in the Experimental group was significantly higher than that of the Control group (P < 0.05). This study indicated that the computer vision detection method could effectively improve the control effect of the carp robot, and had effectiveness and practicability.