Research On Robot Of Packing System Of Elevator Landing Door Based On Machine Vision

With the development of economic,the use of elevators is becoming more and more extensive,the elevator landing door packing is the key link which is the assembly,logistics and distribution of landing door's,its completion is not only related to the quality of the product,but also affected the productivity of the entire production line.In the traditional production line,the landing door is boxed by workers,this way is not only inefficient,but also increase the labour intensity of workers,and even threaten the safety of workers,so the vertical packing system of elevator landing door based on industrial robot is created.This system uses the industrial robot instead of the workers to realize the vertical packing of the landing door,which not only reduce labour intensity,but also improves the working efficiency and realizes the automation of the landing door packing.However,we found some problems through long-term production.Which the robot is used the teaching method,it can not identify error of the box size and positioning,it often appeared in the box door down or the packing door hit a last one in the box.In order to make the door packing more accurately,the system is more intelligent,this paper introduces machine vision technology based on vertical packing system of elevator landing door.The main contents are as follows:First of all,this paper introduces and analyses the relevant technical parameters of the landing door and the boxes,the vertical packing system of elevator landing door based on industrial robot.This system consists of four parts,they are positioning conveyor of case,flipping mechanism of landing door and packing robot,it introduces their respective mechanical structure and the control system.Second,this paper introduces the calibration of visual system and the technology for image processing of landing door.In the case of lens distortion,the ROS system was used to calibrate the camera with the ZHANG method,then solved the internal and external parameters.The image is pre-processed,and finally get a binary image,the use of ROI technology to extract area of the landing door,and then find the number of black and white pixels in the image,and then judge the number of landing door which is packed in the box,and finally the use of Harris method to extract the feature points of the landing door.Thirdly,the SRA166 robot is subjected the kinematics analysis.The kinematics modelling of the robot is established by D-H method,the positive and negative solution of the kinematics is obtained.The robot is simulated and analysed by MATLAB software.Finally,we conduct the experiment of on robot of packing system of elevator landing door based on machine vision,the experiment was carried out to verify the feasibility.For a specific type of landing doors and boxes for packing experiment.The experiment showed that the recognition of the corner of landing door is accurate,the position of position is correct,the relevant issues are resolved,achieve the expected goal.