The Design And Realization Of Household Automatic Massage Robot

With the aggravation of the aging population,the increase of the proportion of the disabled and the rise of the population with chronic diseases,people’s muscle soreness symptoms become more and more daily,and even muscle atrophy and paralysis may occur in severe cases.Muscle soreness requires long-term treatment,not only does it take massive time and vigor for treatment,but also cost is relatively high.Aiming at the above problems,the project mainly studies target location technology and three-dimensional positioning technology in order to design home electronic automatic massage robot.The robot system is developed and designed on the intelligent wheelchair platform.The intelligent wheelchair has the functions of path planning,obstacle avoidance and environment perception,which can solve the problem of robot reaching near the human body.The specific research is as follows:As the beginning,the massage area is recognized by improving human key point detection algorithm of OpenPose.The original OpenPose algorithm consumes a lot of computation,and it will consume substantial memory when deployed in the intelligent wheelchair.Therefore,the direction of the project is to balance the accuracy of the model and the computational cost of the algorithm to satisfy the actual requirements.The network structure of OpenPose algorithm is divided into two modules: feature extraction and detection optimization.In this paper,the vgg-19 used in the original algorithm is replaced by the lighter network module Mobile Net V1,which uses the deep separable convolution algorithm of Mobile Net V1 replaces the convolution algorithm in the original network,and then simplifies the network structure of OpenPose multi-stage detection optimization,only retains the single-stage detection optimization network,and uses multiple small convolution cores to replace the large convolution core in the original algorithm,which greatly reduces the computational complexity.When the computational cost of the algorithm is reduced by about 90% compared with the original algorithm,the accuracy of the algorithm is reduced by only 6%,and its accuracy can still meet the application requirements.Then,using binocular camera,according to the principle of binocular vision,the epipolar constraint method is used to narrow the scope of space search to improve the efficiency of stereo matching,and the orientation constraint algorithm is added to reduce the error of stereo matching.The center of the massage area detected is taken as the massage point to complete the three-dimensional positioning of the massage area.The experimental results show that the accuracy rate reaches 87%,which can meet the requirements This paper introduces the requirements of three-dimensional positioning for home automatic massage robot.Finally,the safety of the home automatic massage robot is analyzed and designed.Combined with the above research content and the actual environment,many experiments are carried out.The results show that the home automatic massage robot designed in this paper can be used stably and safely,and can basically meet the daily needs of users.