Research On Visual Guidance Methods For Industrial Robots

Industrial robots, a key equipment in intelligent manufacturing technologies, has been widely used in all kinds of domains such as automobiles, machinery, electronics, home appliances, petrochemicals and logistics, to implement carrying workpieces, loading and unloading materials, sorting, etc. However, the traditional teaching robots have strong limitations, because they have to finish precise teaching point-by-point before completion of these operations, and can only aim at a single type of workpieces whose positions are fixed. With society developing and industrial demands diversifying, the industrial fields set higher requirements for flexible and intelligent industrial automation production lines. What's more, in recent years the machine vision is developing rapidly, and the robots are beginning to use vision information is beginning to be used to guide the industrial robots to grab the target automatically. Visual guidance technology for industrial robots can make the manufacturing and assembling fully automated, so as to improve production efficiency, therefore it has become a hot research topic nowadays.This thesis mainly studies the key technologies of visual guidance based on a common industrial robot which is six degree of freedom, and combines the visual guidance technologies with the robotic kinematic control to let the robot identify, locate, grab and carry target object automatically. The main research contents and achievements of this paper are follows:1. Kinematic equations of the six degree of freedom industrial robot EPSON C4 is established based on the D-H parameters model, and the tool frame calibration techniques and camera calibration techniques are analyzed and discussed in theories. The calibration on two types of hand-eye systems, Eye-in-Hand system and Eye-to-Hand system, are analyzed with visual guidance technologies. Then their features and applicable cases are summarized.2. An improved multi-resolution SEA template matching algorithm is proposed, which is aimed at workpieces identifying and locating in visual guidance, this improved algorithm can reduce the computation of template matching obviously on the premise of good accuracy. For frequent image interpolation algorithm in template matching, an improved interpolating method is proposed which linearly amplifies image first then interpolates using nearest neighbor algorithm, this improved algorithm can reduce the computation obviously on the premise of interpolating results similar to its of bilinear interpolation algorithm. The experiments results show the effectiveness and computation advantages of the proposed algorithms.3. An experiment platform is built, which is composed of a six degree of freedom industrial robot EPSON C4, an industrial camera, a workbench, and several different shapes of workpieces, for visual guidance robot to carry and assemble. Tool frame calibration, camera calibration and hand-eye system calibration are implemented through software programming in PC and robot controller. The experimental results verify the correctness and feasibility of the visual guidance robot in carrying and assembling program designed by this thesis.