| With the aging of the population,the demand for rehabilitation has also increased sharply,and the existing domestic rehabilitation resources are far from being satisfied.Among them,patients with lower extremity dysfunction need a lot of manpower for nursing and care during the rehabilitation process.The medical follow-up hanger designed in this paper is a medical rehabilitation robot used in this use environment.It can play a protective and supportive role in the patient’s walking rehabilitation training.Patients can use the device to perform walking training independently,reducing the physical labor intensity of medical staff and liberating medical resources during the rehabilitation process.During the recovery process of the patient using this follow-up hanger,just wear the strap and select the automatic follow-up mode,and the hanger can start to work.The patient walks according to their own wishes,and the hanger can detect the patient’s target and follow it for protection.The main research contents of this thesis are as follows:First,the overall design of the mechanical structure and software and hardware system of the medical follower hanger is completed.First design the overall size of the hanger,the suspension structure and the chassis Mecanum wheel structure;Then design the upper-layer hardware system composed of Xavier NX embedded controller and binocular camera and the bottom-layer hardware system with STM32 as the main controller;Install the software environment and design software modules for the hardware system;through the overall construction of the hanger system,it provides a platform basis for the tracking control method.Secondly,the research on human lower limb detection algorithm and binocular ranging algorithm based on YOLOv5 is carried out.The principle of YOLOv5 algorithm and the improvement compared with other versions of the algorithm are introduced.According to the target perspective in the application scenario of the medical follow-up hanger,a data set is created and the model is trained for the YOLOv5 algorithm to obtain good detection results.At the same time,the binocular stereo vision ranging algorithm applied to the medical follower hanger is studied.First,calibrate the binocular camera to reduce the influence of distortion.Then the image is stereo corrected to obtain a line-aligned image,and then matching and depth map hole filling are performed to obtain the actual distance of the target relative to the hanger.After that,the accuracy of binocular ranging is verified,a ranging experiment is designed,and more accurate distance data is obtained.Then,the research of tracking motion and chassis control method-is carried out.According to the typical movements of the lower limbs in the hanger and their pixel positions in the visual image of the hanger,the characteristics of the target frame and the target distance,the tracking motion method was formulated during the patient’s walking training.The overall tracking motion information of the hanger is given,and the rotational speed information of the Mecanum wheel of the chassis is calculated to control the hanger to move with the patient.All of the above design studies have verified the feasibility through separate experiments.Finally,experimental tests are carried out on the hanger to verify the feasibility and overall reliability of the tracking control method.The medical follow-up hanger can realize the followup protection function for patients with lower extremity dysfunction in the process of walking training,liberate medical staff from heavy nursing work,and has broad application prospects. |
