Research On Facial Acupuncture Point Finding And Trajectory Planning Of Acupuncture Robot Based On Binocular Vision

Acupuncture is an important part of traditional Chinese medicine,and after thousands of years of development,it is widely used to treat many diseases around the world;however,acupuncture treatment requires doctors with rich clinical experience,and at the same time,due to the differences between individual patients,there is a large subjectivity and uncertainty in the positioning of acupuncture points,which affects the treatment effect to a certain extent;and at present,the ratio of Chinese doctors to patients in the clinical field of acupuncture is seriously imbalanced.The work of doctors is more intensive.Therefore,in order to solve the above problems and promote the modernization and standardization of traditional acupuncture treatment,this paper completes the localization of human facial acupuncture points based on binocular vision technology,and carries out the planning of acupuncture trajectory for acupuncture robot to complete the acupuncture operation,the study consists of the following components:First,the principle of binocular stereo vision system is analyzed,the conversion of each coordinate system under the system is completed,and the specific process of applying binocular vision technology for 3D reconstruction is clarified.According to the principle of binocular camera calibration and stereo correction,the theoretical method of camera calibration and image correction is determined.The AD-Census stereo matching algorithm was improved to match the left and right images of the patient’s face captured by the binocular camera to obtain accurate parallax information,and based on which the 3D point cloud reconstruction of the human facial model was completed.Secondly,the shape model of ASM algorithm was established to search feature points for the left and right images of the patient’s face and obtain the twodimensional coordinates of feature points.Based on the relevant theory in traditional Chinese medicine of acupuncture,the relative position relationship between facial feature points and acupuncture points is used to determine the facial acupuncture point localization method,obtain the 2D coordinates of acupuncture points in the left and right images of the face,and calculate the corresponding 3D coordinates of each acupuncture point in the facial point cloud model based on the principle of binocular vision.Then,a D-H coordinate system is established for the UR5 e robot used for acupuncture,and robot kinematics and workspace analysis are performed to establish the transformation relationship between the position,posture and each joint angle at the end of the robot.A robot acupuncture trajectory planning algorithm based on NURBS curves is proposed.By completing the estimation of normal vectors from the pre-processed facial point cloud model,and calculating its corresponding robot preparatory acupuncture points by combining the acupuncture posture requirements of relevant acupoints,and completing curve fitting and interpolation based on this,a continuous and smooth robot motion trajectory is established.Finally,the experimental verification of the facial acupuncture of the acupuncture robot was carried out,and the accuracy check was completed for the facial acupuncture point localization method;MATLAB software was used to simulate and analyze the robot acupuncture trajectory to verify the smoothness and fitting effect of the motion trajectory,and a robot facial acupuncture experimental platform was built to control the robot for acupuncture operation by combining the kinematic analysis results and analyzing the accuracy of the actual acupuncture position.The experiments show that the robot facial acupuncture point finding and acupuncture trajectory planning method proposed in this paper can complete the acupuncture operation accurately and efficiently.