Design And Realization Of Industrial Robot Safety System Based On Machine Vision

With the development of our country's economy and the upgrading of industrial production models,as well as the “made in China 2025” strategy,more and more industrial robots play an important role in the field of Manufacturing industry,and it is gradually applied in other fields.The widespread use of industrial robots and the resulting safety incidents are also increasing.It is of great practical significance to solve the safety protection issues of industrial robots.The safety protection methods for traditional industrial robot mainly including “safety cages”,fences,and light curtains,which occupy a large space and have poor flexibility.This thesis proposes the use of machine vision-related technologies to design safety protective system to achieve automatic alarm,remote monitoring,and intelligent processing.The system has good human-computer interaction,can adjust the monitoring area in real time online,and achieve regional classification alarm.At the same time,the system can also intelligently control the robot according to the site conditions to ensure that the robot will not harm the human body.Based on the demand analysis of industrial robot safety protective system and the research of machine vision technology.The thesis proposes the design of safety protective system for industrial robot based on machine vision.In order to solve the problem of the small field of view of ordinary cameras,the thesis is propose to use CCD fisheye camera to capture the image signal of the monitoring area.Image correction is accomplished using fisheye correction method based on latitude and longitude mapping method.For the problem that the fisheye correction algorithm has a large amount of computation and takes a long time,this thesis proposes a scheme to establish a mapping table according to the mapping relationship between the distorted image and the actual position.The image correction will completed through the coordinate table look-up conversion.The thesis analyzes and compares the moving object detection algorithms such as optical flow method,frame difference method and background difference method.It focuses on the background difference method and improves the background modeling method based on the mean method.The thesis uses the projection histogram method to design the target location algorithm and completes the positioning of the moving target.The thesis proposes a regional hierarchical alarm mechanism and alarm control strategy.The thesis works on the embedded DSP platform to achieve the regional settings,fisheye correction,moving target detection and positioning,alarm display and so on.The thesis works has tested on the practical training platform of THMSIR-1 industrial robot.The test resulted show that the thesis completed the design of industrial robot safety protective system based on machine vision and achieved the desired effect.When the personnel break in,the system can realize the regional classification alarm and meet the basic needs of robot safety protection.