Design of Robotic Arm Tracking Based on Face Detection

In recent years, industrial automation technology has developed rapidly, and intelligent robotic arms are widely used in various fields. Robotic arm can help humans to do a variety of work, to a certain extent, to reduce the burden of humans and improve efficiency. With the development of technology, the rapid development of medical-assisted robotic arm, but the current application of medical-assisted robotic arm in surgery still has many limitations, and its problems are mainly to assist in medical work when the trajectory of robotic arm movement and the expected value of the gap, and in the process of movement is not very smooth. On the other hand, the patient cannot maintain a smooth state for a long time during facial treatment, which puts higher requirements on the medical assistance robotic arm, so it is urgent to develop a more accurate medical assistance robotic arm with visual tracking for the above problems. In this paper, we firstly analyze the kinematics of the robotic arm by building a model [1], use polynomial interpolation algorithm to plan the motion trajectory of the robotic arm, and verify the feasibility of the algorithm on Matlab and the physical robotic arm simultaneously, then add a camera module to the robotic arm and train a neural network model to detect the position of the human face. Finally, the face tracking module is designed so that the paramedical robotic arm can follow the face in real time, solving the treatment problem caused by medical patients who cannot maintain a smooth state for a long time when performing facial treatment.