Power-assisted Mobile Robot Based On Deep Vidusl Feedback

As the ageing of the population is getting worse and nurses are getting more and more nervous,helping robots is particularly important.In order to better provide elderly and disabled people with better assistance services,many helper robots have been developed at home and abroad.In view of this,this article has developed a set of assisted mobile robots based on depth visual feedback to achieve some basic auxiliary work in daily life.The main research content of this paper is to transmit the motion information of the human skeleton's feature points captured by the camera to the upper computer.The acquired skeletal point information is processed and recognized by the upper computer,the corresponding control algorithm is designed to realize the human skeletal point and the mapping relationship of the movement robot platform according to the classified processing information,and a smooth trajectory control algorithm is proposed to control the mobile robot.Firstly,a mobile robot motion platform is built by this paper based on depth vision.The depth visual sensor Kinect is used to collect human skeleton information,and the collected motion information is transmitted to the mobile robot platform after being processed by the upper computer so as to realize that the robot is controlled by the human body with the method of modeless control.Because depth-based feedback is used to control mobile robots in this paper,visual calibration is a key step in visual acquisition.Its purpose is to establish the relevant coordinate systems,and the correspondence of spatial points in image coordinate system and space coordinate system obtained by a certain algorithm.Visual calibration is the premise of robot sorting system item detection and capture targets,providing the basis for subsequent crawling.In this paper,the"Zhang Zhengyou" algorithm is used to calibrate the Kinect color camera.Second,this paper proposes a "humanoid joint angle" algorithm to calculate the joint angles of the human upper limb joints.Based on the above research contents,two control softwares,point control software and humanoid control software,were developed.After the development of the humanoid control system,the skeletal points are filtered and the control strategy is optimized for skeletal point fluctuations and complex control problems.After the optimization,the mobile power robot system is utilized for the multiple simulation experiments in a real environment.During the experiment,we observed the motion stability of the mobile platform AGV and the accuracy of the crawling process.After many experiments,we can see that we can achieve the goal of helping the elderly to seize the needed items into the hands of the elderly well and achieve the research purpose of the subject.