Study On Key Technology Of AGV Based On Machine Vision

Automated Guided Vehicle is a pivotal device in Automation Goods Transportation System and Flexible Manufacture Organization System,which can be assembled in any manufacturing system.Meanwhile,it is very easy to transfer parking location and moving path.It converts existing manufacture mode into flexible manufacture organization mode. The degree of industry manufacture automation is advanced.Therefore,high development of AGV is inevitable trend in the industry manufacture area.Image processing technology is the key technology of vision navigated AGV.Aimed at industry actual application AGV based on machinery vision,this article researches on the mark recognition technology of Vision Navigated AGV.It is used to increase the reliability and real-time capability of the image recognition of vision navigated AGV.The research content of this article mainly included:1.Optimization of Splitting Point of Steering Linkage.The principles of kinematics of space mechanism are applied to the analysis of the special motion of strut-link type suspension.The optimization function of the splitting point of Ackerman Steering Linkage is confirmed.The optimal position of Splitting Point is found out,then the swinging angle between wheel and pin is least.2.Image Preprocessing.Image preprocessing includes image strengthening and image segmentation and so on.The purpose of image strengthening is for eliminating the interferential noise in the process of input,transmission and disposal and enhancing the quality of image.Image segmentation is to divide the image into some specifically area with particular character and extract the interested object for object analysis and recognition.3.Self-calibration of camera.Square template is used to calibrate a fixed camera for acquiring the relations between space points and pixel points in the computer picture which is based on circular points.This method requires camera to take the template picture from three different positions.The four edges of the square are detected.At the same time four comer points and midpoints of four edges are extracted.Through circular points calculation,the camera's intrinsic parameters can be obtained linearly,and its extrinsic parameters can also be calculated.4.Feature extraction of mark.Nowadays,the information of some image features including contour features,strip features and gayness statistic features and so on.For the effectively recognition,this article studied the feature matching of numeric numbers recognition based on combined feature.This method can avoid the image processing process including regulation,thinning and contour extraction and so on.5.Mark recognition.Vision navigated AGV has many work conditions in actual application,such as acceleration,deceleration,stop and constant velocity and so on.The intent of designing marks is used to recognize the industry goods transportation destination. According to technical flow,there are many loading and unloading procedure.For distinction different technical procedure,numeric numbers are used to mark procedure.6.The design of the long-distance information display system of AGV based on machinery vision.According to the display requirement of vision navigated AGV,the LCD display interface has designed.The pictures have been transferred to the FLASH ROM of LCD.Procedure has been compiled for display the velocity,direction and pile parameter of AGV in the known coordinates position.These parameters are incepted and sent though the serial port in order to display the long-distance information of vision navigated AGV.